ICLR2025-Papers-with-Code

KO | EN

gitlite — search

#javascript #ai #rust #docker #java #windows #cli #kubernetes #awesome-list #vue #pytorch #ios #openai #open-source

ICLR2025-Papers-with-Code

(★ 590)

历年ICLR论文和开源项目合集，包含ICLR2021、ICLR2022、ICLR2023、ICLR2024、ICLR2025.

ICLR2025-Papers-with-Code Latest Version Download

Download Latest Version (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# ICLR 2025 论文和开源项目合集

>本仓库旨在收集ICLR最新研究进展，尤其是LLM方面，涉及NLP领域的各个方向，此项目长期不定时更新。
>欢迎watch和fork！不过给个star⭐就更好了❤️。
>知乎地址：[**ShuYini**](https://www.zhihu.com/people/wangjini521/activities)
>微信公众号: [**AINLPer**（**每日更新，欢迎关注**）](https://mp.weixin.qq.com/s?__biz=MzUzOTgwNDMzOQ==&mid=2247493926&idx=1&sn=7c3af9a58296f10173b3eba3bec15137&chksm=fac07efacdb7f7ec6f879dad43d5c3dd9af8950ba65faa3cea3f8cbef8208adae2449ec29327&token=890750657&lang=zh_CN#rd)

>另外也欢迎大家进入AINLPer星球，每天推送最新、最优质论文，紧跟AIGC大模型前沿进展；另外星球也特设大模型Agent、大模型推理、RAG系统搭建、热门综述、大模型实操、数据集、测试基准、行业发展状况、大厂工作内推等专栏。详细介绍→：https://mp.weixin.qq.com/s/wHnm9ek4ojYTA_2EPLNILw

>感兴趣的小伙伴，赶快扫描下方⬇二维码，新用户加入更有**50元**优惠券🔖，每年仅需**49**！
>
><img src="https://gitee.com/ainlper/typorapicgo/raw/master/img/image-20250303223035800.png" alt="image-20250303223035800" style="zoom: 50%;" />
>
> 💎💎💎💎💎💎💎💎💎💎💎💎💎💎

## 历年ICLR 论文和开源项目合集

- [**ICLR 2024 论文和开源项目合集**](./ICLR2024-Papers-with-Code.md)
- [**ICLR 2023 论文和开源项目合集**](./ICLR2023-Papers-with-Code.md)
- [**ICLR 2022 论文和开源项目合集**](./ICLR2022-Papers-with-Code.md)
- [**ICLR 2021 论文和开源项目合集**](./ICLR2021-Papers-with-Code.md)

### Accept Spotlight

1、How new data permeates LLM knowledge and how to dilute it

- Paper: [**点击下载**](https://openreview.net/attachment?id=NGKQoaqLpo&name=pdf)
- Code: [https://sunchipsster1.github.io/projects/outlandish/](https://sunchipsster1.github.io/projects/outlandish/)

2、JudgeLM: Fine-tuned Large Language Models are Scalable Judges

- Paper: [**点击下载**](https://openreview.net/attachment?id=xsELpEPn4A&name=pdf)
- Code: [https://github.com/baaivision/JudgeLM](https://github.com/baaivision/JudgeLM)

3、MQuAKE-Remastered: Multi-Hop Knowledge Editing Can Only Be Advanced with Reliable Evaluations

- Paper: [**点击下载**](https://openreview.net/attachment?id=m9wG6ai2Xk&name=pdf)
- Code: [huggingface.co/datasets/henryzhongsc/MQuAKE-Remastered](huggingface.co/datasets/henryzhongsc/MQuAKE-Remastered)

4、Physics of Language Models: Part 3.3, Knowledge Capacity Scaling Laws

- Paper: [**点击下载**](https://openreview.net/attachment?id=FxNNiUgtfa&name=pdf)
- Code: [https://youtu.be/yBL7J0kgldU&t=2220](https://youtu.be/yBL7J0kgldU&t=2220)

5、Internet of Agents: Weaving a Web of Heterogeneous Agents for Collaborative Intelligence

- Paper: [**点击下载**](https://openreview.net/attachment?id=o1Et3MogPw&name=pdf)
- Code: [https://github.com/OpenBMB/IoA](https://github.com/OpenBMB/IoA)

6、Robustness Reprogramming for Representation Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=SuH5SdOXpe&name=pdf)
- Code: [https://github.com/chris-hzc/Robustness-Reprogramming](https://github.com/chris-hzc/Robustness-Reprogramming)

7、LayerDAG: A Layerwise Autoregressive Diffusion Model for Directed Acyclic Graph Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=kam84eEmub&name=pdf)
- Code: [https://github.com/Graph-COM/LayerDAG](https://github.com/Graph-COM/LayerDAG)

8、DailyDilemmas: Revealing Value Preferences of LLMs with Quandaries of Daily Life

- Paper: [**点击下载**](https://openreview.net/attachment?id=PGhiPGBf47&name=pdf)
- Code: [https://hf.co/datasets/kellycyy/daily_dilemmas](https://hf.co/datasets/kellycyy/daily_dilemmas)

9、Learning-Augmented Frequent Directions

- Paper: [**点击下载**](https://openreview.net/attachment?id=WcZLG8XxhD&name=pdf)
- Code: [https://algorithms-with-predictions.github.io/](https://algorithms-with-predictions.github.io/)

10、Quality Measures for Dynamic Graph Generative Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=8bjspmAMBk&name=pdf)
- Code: [https://github.com/ryienh/jl-metric](https://github.com/ryienh/jl-metric)

11、Holistically Evaluating the Environmental Impact of Creating Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=04qx93Viwj&name=pdf)
- Code: [https://www.technologyreview.com/2024/09/26/1104516/three-mile-island-microsoft/](https://www.technologyreview.com/2024/09/26/1104516/three-mile-island-microsoft/)

12、Mixture-of-Agents Enhances Large Language Model Capabilities

- Paper: [**点击下载**](https://openreview.net/attachment?id=h0ZfDIrj7T&name=pdf)
- Code: [https://github.com/togethercomputer/moa](https://github.com/togethercomputer/moa)

13、Answer, Assemble, Ace: Understanding How LMs Answer Multiple Choice Questions

- Paper: [**点击下载**](https://openreview.net/attachment?id=6NNA0MxhCH&name=pdf)
- Code: [https://github.com/allenai/understanding_mcqa](https://github.com/allenai/understanding_mcqa)

14、TopoNets: High performing vision and language models with brain-like topography

- Paper: [**点击下载**](https://openreview.net/attachment?id=THqWPzL00e&name=pdf)
- Code: [https://toponets.github.io](https://toponets.github.io)

15、INCLUDE: Evaluating Multilingual Language Understanding with Regional Knowledge

- Paper: [**点击下载**](https://openreview.net/attachment?id=k3gCieTXeY&name=pdf)
- Code: [https://huggingface.co/datasets/CohereForAI/include-base-44](https://huggingface.co/datasets/CohereForAI/include-base-44)

16、Vision Language Models are In-Context Value Learners

- Paper: [**点击下载**](https://openreview.net/attachment?id=friHAl5ofG&name=pdf)
- Code: [http://generative-value-learning.github.io](http://generative-value-learning.github.io)

17、Wasserstein Distances, Neuronal Entanglement, and Sparsity

- Paper: [**点击下载**](https://openreview.net/attachment?id=cnKhHxN3xj&name=pdf)
- Code: [https://github.com/Shavit-Lab/Sparse-Expansion](https://github.com/Shavit-Lab/Sparse-Expansion)

18、WildBench: Benchmarking LLMs with Challenging Tasks from Real Users in the Wild

- Paper: [**点击下载**](https://openreview.net/attachment?id=MKEHCx25xp&name=pdf)
- Code: [https://hf.co/spaces/allenai/WildBench](https://hf.co/spaces/allenai/WildBench)

19、Adversarial Perturbations Cannot Reliably Protect Artists From Generative AI

- Paper: [**点击下载**](https://openreview.net/attachment?id=yfW1x7uBS5&name=pdf)
- Code: [https://github.com/ethz-spylab/robust-style-mimicry](https://github.com/ethz-spylab/robust-style-mimicry)

20、Multi-session, multi-task neural decoding from distinct cell-types and brain regions

- Paper: [**点击下载**](https://openreview.net/attachment?id=IuU0wcO0mo&name=pdf)
- Code: [https://poyo-plus.github.io](https://poyo-plus.github.io)

21、Reducing Hallucinations in Large Vision-Language Models via Latent Space Steering

- Paper: [**点击下载**](https://openreview.net/attachment?id=LBl7Hez0fF&name=pdf)
- Code: [https://github.com/shengliu66/VTI](https://github.com/shengliu66/VTI)

22、AutoDAN-Turbo: A Lifelong Agent for Strategy Self-Exploration to Jailbreak LLMs

- Paper: [**点击下载**](https://openreview.net/attachment?id=bhK7U37VW8&name=pdf)
- Code: [https://autodans.github.io/AutoDAN-Turbo](https://autodans.github.io/AutoDAN-Turbo)

23、BRIGHT: A Realistic and Challenging Benchmark for Reasoning-Intensive Retrieval

- Paper: [**点击下载**](https://openreview.net/attachment?id=ykuc5q381b&name=pdf)
- Code: [https://hf.co/spaces/mteb/leaderboard?task=retrieval](https://hf.co/spaces/mteb/leaderboard?task=retrieval)

24、Online Reinforcement Learning in Non-Stationary Context-Driven Environments

- Paper: [**点击下载**](https://openreview.net/attachment?id=l6QnSQizmN&name=pdf)
- Code: [https://github.com/pouyahmdn/LCPO](https://github.com/pouyahmdn/LCPO)

25、CEB: Compositional Evaluation Benchmark for Fairness in Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=IUmj2dw5se&name=pdf)
- Code: [https://github.com/SongW-SW/CEB](https://github.com/SongW-SW/CEB)

26、TabWak: A Watermark for Tabular Diffusion Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=71pur4y8gs&name=pdf)
- Code: [https://github.com/chaoyitud/TabWak](https://github.com/chaoyitud/TabWak)

27、Generating Freeform Endoskeletal Robots

- Paper: [**点击下载**](https://openreview.net/attachment?id=awvJBtB2op&name=pdf)
- Code: [https://endoskeletal.github.io](https://endoskeletal.github.io)

28、AnalogGenie: A Generative Engine for Automatic Discovery of Analog Circuit Topologies

- Paper: [**点击下载**](https://openreview.net/attachment?id=jCPak79Kev&name=pdf)
- Code: [https://github.com/xz-group/AnalogGenie](https://github.com/xz-group/AnalogGenie)

29、DataEnvGym: Data Generation Agents in Teacher Environments with Student Feedback

- Paper: [**点击下载**](https://openreview.net/attachment?id=00SnKBGTsz&name=pdf)
- Code: [https://DataEnvGym.github.io](https://DataEnvGym.github.io)

30、LoRA3D: Low-Rank Self-Calibration of 3D Geometric Foundation models

- Paper: [**点击下载**](https://openreview.net/attachment?id=LSp4KBhAom&name=pdf)
- Code: [https://520xyxyzq.github.io/lora3d/](https://520xyxyzq.github.io/lora3d/)

31、Exact Certification of (Graph) Neural Networks Against Label Poisoning

- Paper: [**点击下载**](https://openreview.net/attachment?id=d9aWa875kj&name=pdf)
- Code: [https://github.com/saper0/qpcert](https://github.com/saper0/qpcert)

32、Test-time Adaptation for Cross-modal Retrieval with Query Shift

- Paper: [**点击下载**](https://openreview.net/attachment?id=BmG88rONaU&name=pdf)
- Code: [https://github.com/XLearning-SCU/2025-ICLR-TCR](https://github.com/XLearning-SCU/2025-ICLR-TCR)

33、In Search of Forgotten Domain Generalization

- Paper: [**点击下载**](https://openreview.net/attachment?id=Fk3eod9aaD&name=pdf)
- Code: [https://brendel-group.github.io/clip-dg/](https://brendel-group.github.io/clip-dg/)

34、TOP-ERL: Transformer-based Off-Policy Episodic Reinforcement Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=N4NhVN30ph&name=pdf)
- Code: [https://github.com/BruceGeLi/TOP_ERL_ICLR25](https://github.com/BruceGeLi/TOP_ERL_ICLR25)

35、Harnessing Diversity for Important Data Selection in Pretraining Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=bMC1t7eLRc&name=pdf)
- Code: [https://anonymous.4open.science/r/Quad/](https://anonymous.4open.science/r/Quad/)

36、CausalRivers - Scaling up benchmarking of causal discovery for real-world time-series

- Paper: [**点击下载**](https://openreview.net/attachment?id=wmV4cIbgl6&name=pdf)
- Code: [https://causalrivers.github.io](https://causalrivers.github.io)

37、Provably Reliable Conformal Prediction Sets in the Presence of Data Poisoning

- Paper: [**点击下载**](https://openreview.net/attachment?id=ofuLWn8DFZ&name=pdf)
- Code: [https://www.cs.cit.tum.de/daml/reliable-conformal-prediction/](https://www.cs.cit.tum.de/daml/reliable-conformal-prediction/)

38、Progressive Compositionality in Text-to-Image Generative Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=S85PP4xjFD&name=pdf)
- Code: [https://evansh666.github.io/EvoGen_Page/](https://evansh666.github.io/EvoGen_Page/)

39、Benchmarking Predictive Coding Networks -- Made Simple

- Paper: [**点击下载**](https://openreview.net/attachment?id=sahQq2sH5x&name=pdf)
- Code: [https://github.com/liukidar/pcx](https://github.com/liukidar/pcx)

40、Can Watermarked LLMs be Identified by Users via Crafted Prompts?

- Paper: [**点击下载**](https://openreview.net/attachment?id=ujpAYpFDEA&name=pdf)
- Code: [https://github.com/THU-BPM/Watermarked_LLM_Identification](https://github.com/THU-BPM/Watermarked_LLM_Identification)

41、On Quantizing Neural Representation for Variable-Rate Video Coding

- Paper: [**点击下载**](https://openreview.net/attachment?id=44cMlQSreK&name=pdf)
- Code: [https://github.com/Eric-qi/NeuroQuant](https://github.com/Eric-qi/NeuroQuant)

42、Attention with Markov: A Curious Case of Single-layer Transformers

- Paper: [**点击下载**](https://openreview.net/attachment?id=SqZ0KY4qBD&name=pdf)
- Code: [https://github.com/Bond1995/Markov](https://github.com/Bond1995/Markov)

43、A Second-Order Perspective on Model Compositionality and Incremental Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=OZVTqoli2N&name=pdf)
- Code: [https://github.com/aimagelab/mammoth](https://github.com/aimagelab/mammoth)

44、Iterative Label Refinement Matters More than Preference Optimization under Weak Supervision

- Paper: [**点击下载**](https://openreview.net/attachment?id=q5EZ7gKcnW&name=pdf)
- Code: [https://github.com/helloelwin/iterative-label-refinement](https://github.com/helloelwin/iterative-label-refinement)

45、EmbedLLM: Learning Compact Representations of Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=Fs9EabmQrJ&name=pdf)
- Code: [https://github.com/richardzhuang0412/EmbedLLM](https://github.com/richardzhuang0412/EmbedLLM)

46、Lean-STaR: Learning to Interleave Thinking and Proving

- Paper: [**点击下载**](https://openreview.net/attachment?id=SOWZ59UyNc&name=pdf)
- Code: [https://leanstar.github.io/](https://leanstar.github.io/)

47、NV-Embed: Improved Techniques for Training LLMs as Generalist Embedding Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=lgsyLSsDRe&name=pdf)
- Code: [https://huggingface.co/nvidia/NV-Embed-v2](https://huggingface.co/nvidia/NV-Embed-v2)

48、DARE the Extreme: Revisiting Delta-Parameter Pruning For Fine-Tuned Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=avSocG0oFA&name=pdf)
- Code: [https://github.com/vengdeng/DAREx.git ](https://github.com/vengdeng/DAREx.git )

49、Representative Guidance: Diffusion Model Sampling with Coherence

- Paper: [**点击下载**](https://openreview.net/attachment?id=gWgaypDBs8&name=pdf)
- Code: [https://github.com/dungdinhanh/rep-guidance](https://github.com/dungdinhanh/rep-guidance)

50、Budgeted Online Continual Learning by Adaptive Layer Freezing and Frequency-based Sampling

- Paper: [**点击下载**](https://openreview.net/attachment?id=dOAkHmsjRX&name=pdf)
- Code: [https://github.com/snumprlab/budgeted-cl](https://github.com/snumprlab/budgeted-cl)

51、InverseBench: Benchmarking Plug-and-Play Diffusion Priors for Inverse Problems in Physical Sciences

- Paper: [**点击下载**](https://openreview.net/attachment?id=U3PBITXNG6&name=pdf)
- Code: [https://devzhk.github.io/InverseBench/](https://devzhk.github.io/InverseBench/)

52、Revisiting Random Walks for Learning on Graphs

- Paper: [**点击下载**](https://openreview.net/attachment?id=SG1R2H3fa1&name=pdf)
- Code: [https://github.com/jw9730/random-walk](https://github.com/jw9730/random-walk)

53、DeFT: Decoding with Flash Tree-attention for Efficient Tree-structured LLM Inference

- Paper: [**点击下载**](https://openreview.net/attachment?id=2c7pfOqu9k&name=pdf)
- Code: [https://github.com/LINs-lab/DeFT](https://github.com/LINs-lab/DeFT)

54、Control-oriented Clustering of Visual Latent Representation

- Paper: [**点击下载**](https://openreview.net/attachment?id=pPQPQ7Yd58&name=pdf)
- Code: [https://computationalrobotics.seas.harvard.edu/ControlOriented_NC](https://computationalrobotics.seas.harvard.edu/ControlOriented_NC)

55、AIR-BENCH 2024: A Safety Benchmark based on Regulation and Policies Specified Risk Categories

- Paper: [**点击下载**](https://openreview.net/attachment?id=UVnD9Ze6mF&name=pdf)
- Code: [https://crfm.stanford.edu/helm/air-bench/latest/](https://crfm.stanford.edu/helm/air-bench/latest/)

56、Not All LLM-Generated Data Are Equal: Rethinking Data Weighting in Text Classification

- Paper: [**点击下载**](https://openreview.net/attachment?id=oI5tZaWkF9&name=pdf)
- Code: [https://github.com/Hsun-Yu/DIMP-Loss](https://github.com/Hsun-Yu/DIMP-Loss)

57、X-ALMA: Plug & Play Modules and Adaptive Rejection for Quality Translation at Scale

- Paper: [**点击下载**](https://openreview.net/attachment?id=csbf1p8xUq&name=pdf)
- Code: [https://github.com/fe1ixxu/ALMA](https://github.com/fe1ixxu/ALMA)

58、Token Statistics Transformer: Linear-Time Attention via Variational Rate Reduction

- Paper: [**点击下载**](https://openreview.net/attachment?id=lXRDQsiP2v&name=pdf)
- Code: [https://github.com/RobinWu218/ToST](https://github.com/RobinWu218/ToST)

59、Bilinear MLPs enable weight-based mechanistic interpretability

- Paper: [**点击下载**](https://openreview.net/attachment?id=gI0kPklUKS&name=pdf)
- Code: [https://github.com/tdooms/bilinear-decomposition](https://github.com/tdooms/bilinear-decomposition)

60、Can Large Language Models Understand Symbolic Graphics Programs?

- Paper: [**点击下载**](https://openreview.net/attachment?id=Yk87CwhBDx&name=pdf)
- Code: [https://sgp-bench.github.io](https://sgp-bench.github.io)

61、AgentTrek: Agent Trajectory Synthesis via Guiding Replay with Web Tutorials

- Paper: [**点击下载**](https://openreview.net/attachment?id=EEgYUccwsV&name=pdf)
- Code: [https://agenttrek.github.io](https://agenttrek.github.io)

62、Advantage-Guided Distillation for Preference Alignment in Small Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=xsx3Fpo3UD&name=pdf)
- Code: [https://github.com/SLIT-AI/ADPA](https://github.com/SLIT-AI/ADPA)

63、Simplifying Deep Temporal Difference Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=7IzeL0kflu&name=pdf)
- Code: [https://github.com/mttga/purejaxql](https://github.com/mttga/purejaxql)

64、SVBench: A Benchmark with Temporal Multi-Turn Dialogues for Streaming Video Understanding

- Paper: [**点击下载**](https://openreview.net/attachment?id=Hz4BYVY8YM&name=pdf)
- Code: [https://yzy-bupt.github.io/SVBench](https://yzy-bupt.github.io/SVBench)

65、Biologically Constrained Barrel Cortex Model Integrates Whisker Inputs and Replicates Key Brain Network Dynamics

- Paper: [**点击下载**](https://openreview.net/attachment?id=UvfI4grcM7&name=pdf)
- Code: [https://github.com/fun0515/RSNN_bfd](https://github.com/fun0515/RSNN_bfd)

66、Broaden your SCOPE! Efficient Multi-turn Conversation Planning for LLMs with Semantic Space

- Paper: [**点击下载**](https://openreview.net/attachment?id=3cgMU3TyyE&name=pdf)
- Code: [https://github.com/chenzhiliang94/convo-plan-SCOPE](https://github.com/chenzhiliang94/convo-plan-SCOPE)

67、The Superposition of Diffusion Models Using the Itô Density Estimator

- Paper: [**点击下载**](https://openreview.net/attachment?id=2o58Mbqkd2&name=pdf)
- Code: [https://github.com/necludov/super-diffusion](https://github.com/necludov/super-diffusion)

68、MAGNet: Motif-Agnostic Generation of Molecules from Scaffolds

- Paper: [**点击下载**](https://openreview.net/attachment?id=5FXKgOxmb2&name=pdf)
- Code: [https://www.cs.cit.tum.de/daml/magnet/](https://www.cs.cit.tum.de/daml/magnet/)

69、Interleaved Scene Graphs for Interleaved Text-and-Image Generation Assessment

- Paper: [**点击下载**](https://openreview.net/attachment?id=rDLgnYLM5b&name=pdf)
- Code: [https://interleave-eval.github.io](https://interleave-eval.github.io)

70、IGL-Bench: Establishing the Comprehensive Benchmark for Imbalanced Graph Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=uTqnyF0JNR&name=pdf)
- Code: [https://github.com/RingBDStack/IGL-Bench](https://github.com/RingBDStack/IGL-Bench)

71、RegMix: Data Mixture as Regression for Language Model Pre-training

- Paper: [**点击下载**](https://openreview.net/attachment?id=5BjQOUXq7i&name=pdf)
- Code: [https://github.com/sail-sg/regmix](https://github.com/sail-sg/regmix)

72、When Attention Sink Emerges in Language Models: An Empirical View

- Paper: [**点击下载**](https://openreview.net/attachment?id=78Nn4QJTEN&name=pdf)
- Code: [https://github.com/sail-sg/Attention-Sink](https://github.com/sail-sg/Attention-Sink)

73、PianoMotion10M: Dataset and Benchmark for Hand Motion Generation in Piano Performance

- Paper: [**点击下载**](https://openreview.net/attachment?id=rxVvRBgqmS&name=pdf)
- Code: [https://github.com/agnJason/PianoMotion10M](https://github.com/agnJason/PianoMotion10M)

74、Revisiting text-to-image evaluation with Gecko: on metrics, prompts, and human rating

- Paper: [**点击下载**](https://openreview.net/attachment?id=Im2neAMlre&name=pdf)
- Code: [https://github.com/google-deepmind/gecko_benchmark_t2i](https://github.com/google-deepmind/gecko_benchmark_t2i)

75、Streamlining Redundant Layers to Compress Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=IC5RJvRoMp&name=pdf)
- Code: [https://github.com/RUCKBReasoning/LLM-Streamline](https://github.com/RUCKBReasoning/LLM-Streamline)

76、SynFlowNet: Design of Diverse and Novel Molecules with Synthesis Constraints

- Paper: [**点击下载**](https://openreview.net/attachment?id=uvHmnahyp1&name=pdf)
- Code: [https://github.com/mirunacrt/synflownet](https://github.com/mirunacrt/synflownet)

77、Rethinking and Improving Autoformalization: Towards a Faithful Metric and a Dependency Retrieval-based Approach

- Paper: [**点击下载**](https://openreview.net/attachment?id=hUb2At2DsQ&name=pdf)
- Code: [https://github.com/Purewhite2019/rethinking_autoformalization](https://github.com/Purewhite2019/rethinking_autoformalization)

78、Learning Spatiotemporal Dynamical Systems from Point Process Observations

- Paper: [**点击下载**](https://openreview.net/attachment?id=37EXtKCOkn&name=pdf)
- Code: [https://github.com/yakovlev31/pproc-dyn](https://github.com/yakovlev31/pproc-dyn)

79、Probabilistic Neural Pruning via Sparsity Evolutionary Fokker-Planck-Kolmogorov Equation

- Paper: [**点击下载**](https://openreview.net/attachment?id=hJ1BaJ5ELp&name=pdf)
- Code: [https://github.com/mzf666/SFPK-main](https://github.com/mzf666/SFPK-main)

80、Uncovering Gaps in How Humans and LLMs Interpret Subjective Language

- Paper: [**点击下载**](https://openreview.net/attachment?id=gye2U9uNXx&name=pdf)
- Code: [https://github.com/arjunpat/thesaurus-error-detector](https://github.com/arjunpat/thesaurus-error-detector)

81、SPA-BENCH: A COMPREHENSIVE BENCHMARK FOR SMARTPHONE AGENT EVALUATION

- Paper: [**点击下载**](https://openreview.net/attachment?id=OZbFRNhpwr&name=pdf)
- Code: [https://ai-agents-2030.github.io/SPA-Bench/](https://ai-agents-2030.github.io/SPA-Bench/)

82、Demystifying the Token Dynamics of Deep Selective State Space Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=qtTIP5Gjc5&name=pdf)
- Code: [https://github.com/Fsoft-AIC/Mamba-token-dynamic](https://github.com/Fsoft-AIC/Mamba-token-dynamic)

83、Knowledge Localization: Mission Not Accomplished? Enter Query Localization!

- Paper: [**点击下载**](https://openreview.net/attachment?id=tfyHbvFZ0K&name=pdf)
- Code: [https://github.com/heng840/KnowledgeLocalization](https://github.com/heng840/KnowledgeLocalization)

84、Realistic Evaluation of Deep Partial-Label Learning Algorithms

- Paper: [**点击下载**](https://openreview.net/attachment?id=FtX6oAW7Dd&name=pdf)
- Code: [https://github.com/wwangwitsel/PLENCH](https://github.com/wwangwitsel/PLENCH)

85、Graph Sparsification via Mixture of Graphs

- Paper: [**点击下载**](https://openreview.net/attachment?id=7ANDviElAo&name=pdf)
- Code: [https://anonymous.4open.science/r/MoG-5E0F](https://anonymous.4open.science/r/MoG-5E0F)

86、RAG-SR: Retrieval-Augmented Generation for Neural Symbolic Regression

- Paper: [**点击下载**](https://openreview.net/attachment?id=NdHka08uWn&name=pdf)
- Code: [https://github.com/hengzhe-zhang/RAG-SR](https://github.com/hengzhe-zhang/RAG-SR)

87、MixEval-X: Any-to-any Evaluations from Real-world Data Mixture

- Paper: [**点击下载**](https://openreview.net/attachment?id=hpCfPEvBsr&name=pdf)
- Code: [https://mixeval-x.github.io/](https://mixeval-x.github.io/)

88、DEEM: Diffusion models serve as the eyes of large language models for image perception

- Paper: [**点击下载**](https://openreview.net/attachment?id=qtWjSboqfe&name=pdf)
- Code: [https://github.com/RainBowLuoCS/DEEM](https://github.com/RainBowLuoCS/DEEM)

89、BodyGen: Advancing Towards Efficient Embodiment Co-Design

- Paper: [**点击下载**](https://openreview.net/attachment?id=cTR17xl89h&name=pdf)
- Code: [https://genesisorigin.github.io](https://genesisorigin.github.io)

90、Diffusion Bridge AutoEncoders for Unsupervised Representation Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=hBGavkf61a&name=pdf)
- Code: [https://github.com/aailab-kaist/DBAE](https://github.com/aailab-kaist/DBAE)

91、Bi-Factorial Preference Optimization: Balancing Safety-Helpfulness in Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=GjM61KRiTG&name=pdf)
- Code: [https://github.com/wx-zhang/bfpo](https://github.com/wx-zhang/bfpo)

92、Grounding Video Models to Actions through Goal Conditioned Exploration

- Paper: [**点击下载**](https://openreview.net/attachment?id=G6dMvRuhFr&name=pdf)
- Code: [https://video-to-action.github.io/](https://video-to-action.github.io/)

93、Dense Video Object Captioning from Disjoint Supervision

- Paper: [**点击下载**](https://openreview.net/attachment?id=auZZ2gN0ZN&name=pdf)
- Code: [https://github.com/google-research/scenic/tree/main/scenic/projects/densevoc](https://github.com/google-research/scenic/tree/main/scenic/projects/densevoc)

94、RESuM: A Rare Event Surrogate Model for  Physics Detector Design

- Paper: [**点击下载**](https://openreview.net/attachment?id=lqTILjL6lP&name=pdf)
- Code: [https://github.com/annkasch/resum-legend](https://github.com/annkasch/resum-legend)

95、DeLLMa: Decision Making Under Uncertainty with Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=Acvo2RGSCy&name=pdf)
- Code: [https://dellma.github.io/](https://dellma.github.io/)

96、Stabilizing Reinforcement Learning in Differentiable Multiphysics Simulation

- Paper: [**点击下载**](https://openreview.net/attachment?id=DRiLWb8bJg&name=pdf)
- Code: [https://rewarped.github.io](https://rewarped.github.io)

97、A Periodic Bayesian Flow for Material Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=Lz0XW99tE0&name=pdf)
- Code: [https://github.com/wu-han-lin/CrysBFN](https://github.com/wu-han-lin/CrysBFN)

98、SplatFormer: Point Transformer for Robust 3D Gaussian Splatting

- Paper: [**点击下载**](https://openreview.net/attachment?id=9NfHbWKqMF&name=pdf)
- Code: [https://sergeyprokudin.github.io/splatformer](https://sergeyprokudin.github.io/splatformer)

99、Continuous Exposure Learning for Low-light Image Enhancement using Neural ODEs

- Paper: [**点击下载**](https://openreview.net/attachment?id=Mn2qgIcIPS&name=pdf)
- Code: [https://github.com/dgjung0220/CLODE](https://github.com/dgjung0220/CLODE)

100、Lumina-T2X: Scalable Flow-based Large Diffusion Transformer for Flexible Resolution Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=EbWf36quzd&name=pdf)
- Code: [https://github.com/Alpha-VLLM/Lumina-T2X](https://github.com/Alpha-VLLM/Lumina-T2X)

101、Towards Marginal Fairness Sliced Wasserstein Barycenter

- Paper: [**点击下载**](https://openreview.net/attachment?id=NQqJPPCesd&name=pdf)
- Code: [https://github.com/khainb/MFSWB](https://github.com/khainb/MFSWB)

102、Co$^{\mathbf{3}}$Gesture: Towards Coherent Concurrent Co-speech 3D Gesture Generation with Interactive Diffusion

- Paper: [**点击下载**](https://openreview.net/attachment?id=VaowElpVzd&name=pdf)
- Code: [https://mattie-e.github.io/Co3/](https://mattie-e.github.io/Co3/)

103、Following the Human Thread in Social Navigation

- Paper: [**点击下载**](https://openreview.net/attachment?id=M8OGl34Pmg&name=pdf)
- Code: [https://github.com/L-Scofano/SDA](https://github.com/L-Scofano/SDA)

104、OS-ATLAS: Foundation Action Model for Generalist GUI Agents

- Paper: [**点击下载**](https://openreview.net/attachment?id=n9PDaFNi8t&name=pdf)
- Code: [https://osatlas.github.io/](https://osatlas.github.io/)

105、Student-Informed Teacher Training

- Paper: [**点击下载**](https://openreview.net/attachment?id=Dzh0hQPpuf&name=pdf)
- Code: [https://rpg.ifi.uzh.ch/sitt/](https://rpg.ifi.uzh.ch/sitt/)

106、Preference Optimization for Reasoning with Pseudo Feedback

- Paper: [**点击下载**](https://openreview.net/attachment?id=jkUp3lybXf&name=pdf)
- Code: [https://github.com/microsoft/unilm/tree/master/PFPO](https://github.com/microsoft/unilm/tree/master/PFPO)

107、OmniRe: Omni Urban Scene Reconstruction

- Paper: [**点击下载**](https://openreview.net/attachment?id=11xgiMEI5o&name=pdf)
- Code: [https://ziyc.github.io/omnire/](https://ziyc.github.io/omnire/)

108、Multimodality Helps Few-shot 3D Point Cloud Semantic Segmentation

- Paper: [**点击下载**](https://openreview.net/attachment?id=jXvwJ51vcK&name=pdf)
- Code: [https://github.com/ZhaochongAn/Multimodality-3D-Few-Shot](https://github.com/ZhaochongAn/Multimodality-3D-Few-Shot)

109、Multi-modal Agent Tuning: Building a VLM-Driven Agent for Efficient Tool Usage

- Paper: [**点击下载**](https://openreview.net/attachment?id=0bmGL4q7vJ&name=pdf)
- Code: [https://mat-agent.github.io](https://mat-agent.github.io)

110、Multi-Field Adaptive Retrieval

- Paper: [**点击下载**](https://openreview.net/attachment?id=3PDklqqqfN&name=pdf)
- Code: [https://github.com/microsoft/multifield-adaptive-retrieval](https://github.com/microsoft/multifield-adaptive-retrieval)

111、On Disentangled Training for Nonlinear Transform in Learned Image Compression

- Paper: [**点击下载**](https://openreview.net/attachment?id=U67J0QNtzo&name=pdf)
- Code: [https://github.com/qingshi9974/AuxT](https://github.com/qingshi9974/AuxT)

112、MMAU: A Massive Multi-Task Audio Understanding and Reasoning Benchmark

- Paper: [**点击下载**](https://openreview.net/attachment?id=TeVAZXr3yv&name=pdf)
- Code: [https://sakshi113.github.io/mmau_homepage/](https://sakshi113.github.io/mmau_homepage/)

113、ReDeEP: Detecting Hallucination in Retrieval-Augmented Generation via Mechanistic Interpretability

- Paper: [**点击下载**](https://openreview.net/attachment?id=ztzZDzgfrh&name=pdf)
- Code: [https://github.com/Jeryi-Sun/ReDEeP-ICLR](https://github.com/Jeryi-Sun/ReDEeP-ICLR)

114、Towards General-Purpose Model-Free Reinforcement Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=R1hIXdST22&name=pdf)
- Code: [https://github.com/facebookresearch/MRQ](https://github.com/facebookresearch/MRQ)

115、Accelerating Goal-Conditioned Reinforcement Learning Algorithms and Research

- Paper: [**点击下载**](https://openreview.net/attachment?id=4gaySj8kvX&name=pdf)
- Code: [https://github.com/MichalBortkiewicz/JaxGCRL](https://github.com/MichalBortkiewicz/JaxGCRL)

116、Time-MoE: Billion-Scale Time Series Foundation Models with Mixture of Experts

- Paper: [**点击下载**](https://openreview.net/attachment?id=e1wDDFmlVu&name=pdf)
- Code: [https://github.com/Time-MoE/Time-MoE](https://github.com/Time-MoE/Time-MoE)

117、SimBa: Simplicity Bias for Scaling Up Parameters in Deep Reinforcement Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=jXLiDKsuDo&name=pdf)
- Code: [https://sonyresearch.github.io/simba](https://sonyresearch.github.io/simba)

118、Self-play with Execution Feedback: Improving Instruction-following Capabilities of Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=cRR0oDFEBC&name=pdf)
- Code: [https://github.com/QwenLM/AutoIF](https://github.com/QwenLM/AutoIF)

119、Diffusion On Syntax Trees For Program Synthesis

- Paper: [**点击下载**](https://openreview.net/attachment?id=wN3KaUXA5X&name=pdf)
- Code: [https://tree-diffusion.github.io](https://tree-diffusion.github.io)

120、Vision-RWKV: Efficient and Scalable Visual Perception with RWKV-Like Architectures

- Paper: [**点击下载**](https://openreview.net/attachment?id=nGiGXLnKhl&name=pdf)
- Code: [https://github.com/OpenGVLab/Vision-RWKV](https://github.com/OpenGVLab/Vision-RWKV)

121、A CLIP-Powered Framework for Robust and Generalizable Data Selection

- Paper: [**点击下载**](https://openreview.net/attachment?id=9bMZ29SPVx&name=pdf)
- Code: [https://github.com/Jackbrocp/clip-powered-data-selection](https://github.com/Jackbrocp/clip-powered-data-selection)

122、Fine-tuning with Reserved Majority for Noise Reduction

- Paper: [**点击下载**](https://openreview.net/attachment?id=ZV7CLf0RHK&name=pdf)
- Code: [https://github.com/pixas/NoRM](https://github.com/pixas/NoRM)

123、GOLD: Graph Out-of-Distribution Detection via Implicit Adversarial Latent Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=y5einmJ0Yx&name=pdf)
- Code: [https://github.com/DannyW618/GOLD](https://github.com/DannyW618/GOLD)

124、Topograph: An Efficient Graph-Based Framework for Strictly Topology Preserving Image Segmentation

- Paper: [**点击下载**](https://openreview.net/attachment?id=Q0zmmNNePz&name=pdf)
- Code: [https://anonymous.4open.science/r/Topograph](https://anonymous.4open.science/r/Topograph)

125、Test-time Alignment of Diffusion Models without Reward Over-optimization

- Paper: [**点击下载**](https://openreview.net/attachment?id=vi3DjUhFVm&name=pdf)
- Code: [https://github.com/krafton-ai/DAS](https://github.com/krafton-ai/DAS)

126、ImpScore: A Learnable Metric For Quantifying The Implicitness Level of Sentences

- Paper: [**点击下载**](https://openreview.net/attachment?id=gYWqxXE5RJ&name=pdf)
- Code: [https://github.com/audreycs/ImpScore](https://github.com/audreycs/ImpScore)

127、LoRA-Pro: Are Low-Rank Adapters Properly Optimized?

- Paper: [**点击下载**](https://openreview.net/attachment?id=gTwRMU3lJ5&name=pdf)
- Code: [https://github.com/mrflogs/LoRA-Pro](https://github.com/mrflogs/LoRA-Pro)

128、LLaVA-NeXT-Interleave: Tackling Multi-image, Video, and 3D in Large Multimodal Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=oSQiao9GqB&name=pdf)
- Code: [https://llava-vl.github.io/blog/2024-06-16-llava-next-interleave/](https://llava-vl.github.io/blog/2024-06-16-llava-next-interleave/)

129、Large-scale and Fine-grained Vision-language Pre-training for Enhanced CT Image Understanding

- Paper: [**点击下载**](https://openreview.net/attachment?id=nYpPAT4L3D&name=pdf)
- Code: [https://github.com/alibaba-damo-academy/fvlm](https://github.com/alibaba-damo-academy/fvlm)

130、GETS: Ensemble Temperature Scaling for Calibration in Graph Neural Networks

- Paper: [**点击下载**](https://openreview.net/attachment?id=qgsXsqahMq&name=pdf)
- Code: [https://github.com/ZhuangDingyi/GETS/](https://github.com/ZhuangDingyi/GETS/)

131、Easing Training Process of Rectified Flow Models Via Lengthening Inter-Path Distance

- Paper: [**点击下载**](https://openreview.net/attachment?id=RaR3ETzyKp&name=pdf)
- Code: [https://github.com/shifengxu/DANSM](https://github.com/shifengxu/DANSM)

132、Recovering Manifold Structure Using Ollivier Ricci Curvature

- Paper: [**点击下载**](https://openreview.net/attachment?id=aX7X9z3vQS&name=pdf)
- Code: [https://github.com/TristanSaidi/orcml](https://github.com/TristanSaidi/orcml)

133、AutoCGP: Closed-Loop Concept-Guided Policies from Unlabeled Demonstrations

- Paper: [**点击下载**](https://openreview.net/attachment?id=9ehJCZz4aM&name=pdf)
- Code: [https://github.com/PeiZhou26/AutoCGP](https://github.com/PeiZhou26/AutoCGP)

134、Min-K%++: Improved Baseline for Pre-Training Data Detection from Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=ZGkfoufDaU&name=pdf)
- Code: [https://zjysteven.github.io/mink-plus-plus/](https://zjysteven.github.io/mink-plus-plus/)

135、MonST3R: A Simple Approach for Estimating Geometry in the Presence of Motion

- Paper: [**点击下载**](https://openreview.net/attachment?id=lJpqxFgWCM&name=pdf)
- Code: [https://monst3r-project.github.io/](https://monst3r-project.github.io/)

136、Atlas Gaussians Diffusion for 3D Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=H2Gxil855b&name=pdf)
- Code: [https://yanghtr.github.io/projects/atlas_gaussians](https://yanghtr.github.io/projects/atlas_gaussians)

137、Samba: Synchronized Set-of-Sequences Modeling for Multiple Object Tracking

- Paper: [**点击下载**](https://openreview.net/attachment?id=OeBY9XqiTz&name=pdf)
- Code: [https://sambamotr.github.io/](https://sambamotr.github.io/)

138、Eagle: Exploring The Design Space for Multimodal LLMs with Mixture of Encoders

- Paper: [**点击下载**](https://openreview.net/attachment?id=Y2RW9EVwhT&name=pdf)
- Code: [https://github.com/NVlabs/Eagle](https://github.com/NVlabs/Eagle)

139、Programming Refusal with Conditional Activation Steering

- Paper: [**点击下载**](https://openreview.net/attachment?id=Oi47wc10sm&name=pdf)
- Code: [https://github.com/IBM/activation-steering](https://github.com/IBM/activation-steering)

140、Perm: A Parametric Representation for Multi-Style 3D Hair Modeling

- Paper: [**点击下载**](https://openreview.net/attachment?id=WKfb1xGXGx&name=pdf)
- Code: [https://cs.yale.edu/homes/che/projects/perm/](https://cs.yale.edu/homes/che/projects/perm/)

141、CLoSD: Closing the Loop between Simulation and Diffusion for multi-task character control

- Paper: [**点击下载**](https://openreview.net/attachment?id=pZISppZSTv&name=pdf)
- Code: [https://guytevet.github.io/CLoSD-page/](https://guytevet.github.io/CLoSD-page/)

142、ND-SDF: Learning Normal Deflection Fields for High-Fidelity Indoor Reconstruction

- Paper: [**点击下载**](https://openreview.net/attachment?id=4HRRcqE9SU&name=pdf)
- Code: [https://zju3dv.github.io/nd-sdf/](https://zju3dv.github.io/nd-sdf/)

143、D-FINE: Redefine Regression Task of DETRs as Fine-grained Distribution Refinement

- Paper: [**点击下载**](https://openreview.net/attachment?id=MFZjrTFE7h&name=pdf)
- Code: [https://github.com/Peterande/D-FINE](https://github.com/Peterande/D-FINE)

144、Mitigating Information Loss in Tree-Based Reinforcement Learning via Direct Optimization

- Paper: [**点击下载**](https://openreview.net/attachment?id=qpXctF2aLZ&name=pdf)
- Code: [https://github.com/s-marton/sympol](https://github.com/s-marton/sympol)

145、Sharpness-Aware Minimization Efficiently Selects Flatter Minima Late In Training

- Paper: [**点击下载**](https://openreview.net/attachment?id=aD2uwhLbnA&name=pdf)
- Code: [https://github.com/zzp1012/SAM-in-Late-Phase](https://github.com/zzp1012/SAM-in-Late-Phase)

146、DartControl: A Diffusion-Based Autoregressive Motion Model for Real-Time Text-Driven Motion Control

- Paper: [**点击下载**](https://openreview.net/attachment?id=XNA3Mnnbvb&name=pdf)
- Code: [https://zkf1997.github.io/DART/](https://zkf1997.github.io/DART/)

147、ThinK: Thinner Key Cache by Query-Driven Pruning

- Paper: [**点击下载**](https://openreview.net/attachment?id=n0OtGl6VGb&name=pdf)
- Code: [https://github.com/SalesforceAIResearch/ThinK](https://github.com/SalesforceAIResearch/ThinK)

148、4K4DGen: Panoramic 4D Generation at 4K Resolution

- Paper: [**点击下载**](https://openreview.net/attachment?id=qxRoo7ULCo&name=pdf)
- Code: [https://4k4dgen.github.io/](https://4k4dgen.github.io/)

149、RelitLRM: Generative Relightable Radiance for Large Reconstruction Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=3Oli4u6q3p&name=pdf)
- Code: [https://relit-lrm.github.io/](https://relit-lrm.github.io/)

150、ConFIG: Towards Conflict-free Training of Physics Informed Neural Networks

- Paper: [**点击下载**](https://openreview.net/attachment?id=APojAzJQiq&name=pdf)
- Code: [https://tum-pbs.github.io/ConFIG](https://tum-pbs.github.io/ConFIG)

151、Severing Spurious Correlations with Data Pruning

- Paper: [**点击下载**](https://openreview.net/attachment?id=Bk13Qfu8Ru&name=pdf)
- Code: [https://github.com/JEKimLab/ICLR2025_SpuriousDataPruning](https://github.com/JEKimLab/ICLR2025_SpuriousDataPruning)

152、MotionAura: Generating High-Quality and Motion Consistent Videos using Discrete Diffusion

- Paper: [**点击下载**](https://openreview.net/attachment?id=bW9fGYo44s&name=pdf)
- Code: [https://researchgroup12.github.io/Abstract_Diagram.html](https://researchgroup12.github.io/Abstract_Diagram.html)

153、Hymba: A Hybrid-head Architecture for Small Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=A1ztozypga&name=pdf)
- Code: [https://huggingface.co/nvidia/Hymba-1.5B-Base](https://huggingface.co/nvidia/Hymba-1.5B-Base)

154、Training-Free Activation Sparsity in Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=dGVZwyq5tV&name=pdf)
- Code: [https://github.com/FasterDecoding/TEAL](https://github.com/FasterDecoding/TEAL)

155、Determine-Then-Ensemble: Necessity of Top-k Union for Large Language Model Ensembling

- Paper: [**点击下载**](https://openreview.net/attachment?id=FDnZFpHmU4&name=pdf)
- Code: [https://github.com/starrYYxuan/UniTE](https://github.com/starrYYxuan/UniTE)

156、CubeDiff: Repurposing Diffusion-Based Image Models for Panorama Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=M2SsqpxGtc&name=pdf)
- Code: [https://cubediff.github.io/](https://cubediff.github.io/)

157、Poison-splat: Computation Cost Attack on 3D Gaussian Splatting

- Paper: [**点击下载**](https://openreview.net/attachment?id=ExrEw8cVlU&name=pdf)
- Code: [ https://github.com/jiahaolu97/poison-splat]( https://github.com/jiahaolu97/poison-splat)

158、Rare-to-Frequent: Unlocking Compositional Generation Power of Diffusion Models on Rare Concepts with LLM Guidance

- Paper: [**点击下载**](https://openreview.net/attachment?id=BgxsmpVoOX&name=pdf)
- Code: [https://github.com/krafton-ai/Rare-to-Frequent](https://github.com/krafton-ai/Rare-to-Frequent)

159、Scalable and Certifiable Graph Unlearning: Overcoming the Approximation Error Barrier

- Paper: [**点击下载**](https://openreview.net/attachment?id=pPyJyeLriR&name=pdf)
- Code: [https://github.com/luyi256/ScaleGUN](https://github.com/luyi256/ScaleGUN)

160、TokenFormer: Rethinking Transformer Scaling with Tokenized Model Parameters

- Paper: [**点击下载**](https://openreview.net/attachment?id=oQ4igHyh3N&name=pdf)
- Code: [https://github.com/Haiyang-W/TokenFormer.git](https://github.com/Haiyang-W/TokenFormer.git)

161、Linear Spherical Sliced Optimal Transport: A Fast Metric for Comparing Spherical Data

- Paper: [**点击下载**](https://openreview.net/attachment?id=fgUFZAxywx&name=pdf)
- Code: [adni.loni.usc.edu](adni.loni.usc.edu)

162、Enhancing Pre-trained Representation Classifiability can Boost its Interpretability

- Paper: [**点击下载**](https://openreview.net/attachment?id=GjfIZan5jN&name=pdf)
- Code: [https://github.com/ssfgunner/IIS](https://github.com/ssfgunner/IIS)

163、LiveBench: A Challenging, Contamination-Limited LLM Benchmark

- Paper: [**点击下载**](https://openreview.net/attachment?id=sKYHBTAxVa&name=pdf)
- Code: [https://livebench.ai](https://livebench.ai)

164、Stem-OB: Generalizable Visual Imitation Learning with Stem-Like Convergent Observation through Diffusion Inversion

- Paper: [**点击下载**](https://openreview.net/attachment?id=xaYlO03tIk&name=pdf)
- Code: [https://github.com/hukz18/Stem-Ob/](https://github.com/hukz18/Stem-Ob/)

165、Reti-Diff: Illumination Degradation Image Restoration with Retinex-based Latent Diffusion Model

- Paper: [**点击下载**](https://openreview.net/attachment?id=kxFtMHItrf&name=pdf)
- Code: [https://github.com/ChunmingHe/Reti-Diff](https://github.com/ChunmingHe/Reti-Diff)

166、MRS: A Fast Sampler for Mean Reverting Diffusion based on ODE and SDE Solvers

- Paper: [**点击下载**](https://openreview.net/attachment?id=yVeNBxwL5W&name=pdf)
- Code: [https://github.com/grrrute/mr-sampler](https://github.com/grrrute/mr-sampler)

167、SRSA: Skill Retrieval and Adaptation for Robotic Assembly Tasks

- Paper: [**点击下载**](https://openreview.net/attachment?id=RInisw1yin&name=pdf)
- Code: [https://srsa2024.github.io/](https://srsa2024.github.io/)

168、DenseMatcher: Learning 3D Semantic Correspondence for Category-Level Manipulation from a Single Demo

- Paper: [**点击下载**](https://openreview.net/attachment?id=8oFvUBvF1u&name=pdf)
- Code: [https://tea-lab.github.io/DenseMatcher/](https://tea-lab.github.io/DenseMatcher/)

169、Beyond Next Token Prediction: Patch-Level Training for Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=dDpB23VbVa&name=pdf)
- Code: [https://github.com/shaochenze/PatchTrain](https://github.com/shaochenze/PatchTrain)

170、LeFusion: Controllable Pathology Synthesis via Lesion-Focused Diffusion Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=3b9SKkRAKw&name=pdf)
- Code: [https://github.com/M3DV/LeFusion](https://github.com/M3DV/LeFusion)

171、LOKI: A Comprehensive Synthetic Data Detection Benchmark using Large Multimodal Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=z8sxoCYgmd&name=pdf)
- Code: [https://opendatalab.github.io/LOKI/](https://opendatalab.github.io/LOKI/)

172、Both Ears Wide Open: Towards Language-Driven Spatial Audio Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=qPx3i9sMxv&name=pdf)
- Code: [https://peiwensun2000.github.io/bewo/](https://peiwensun2000.github.io/bewo/)

173、One-Prompt-One-Story: Free-Lunch Consistent Text-to-Image Generation Using a Single Prompt

- Paper: [**点击下载**](https://openreview.net/attachment?id=cD1kl2QKv1&name=pdf)
- Code: [https://github.com/byliutao/1Prompt1Story](https://github.com/byliutao/1Prompt1Story)

174、MetaUrban: An Embodied AI Simulation Platform for Urban Micromobility

- Paper: [**点击下载**](https://openreview.net/attachment?id=kFsWpSxkFz&name=pdf)
- Code: [https://metadriverse.github.io/metaurban](https://metadriverse.github.io/metaurban)

175、UniMatch: Universal Matching from Atom to Task for Few-Shot Drug Discovery

- Paper: [**点击下载**](https://openreview.net/attachment?id=v9EjwMM55Y&name=pdf)
- Code: [https://github.com/Lirain21/UniMatch.git](https://github.com/Lirain21/UniMatch.git)

176、DynamicCity: Large-Scale 4D Occupancy Generation from Dynamic Scenes

- Paper: [**点击下载**](https://openreview.net/attachment?id=M7KyLjuN0A&name=pdf)
- Code: [https://dynamic-city.github.io](https://dynamic-city.github.io)

177、Instance-dependent Early Stopping

- Paper: [**点击下载**](https://openreview.net/attachment?id=P42DbV2nuV&name=pdf)
- Code: [https://github.com/tmllab/2025_ICLR_IES](https://github.com/tmllab/2025_ICLR_IES)

178、Moner: Motion Correction in Undersampled Radial MRI with Unsupervised Neural Representation

- Paper: [**点击下载**](https://openreview.net/attachment?id=OdnqG1fYpo&name=pdf)
- Code: [https://github.com/iwuqing/Moner](https://github.com/iwuqing/Moner)

179、SVDQuant: Absorbing Outliers by Low-Rank Component for 4-Bit Diffusion Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=vWR3KuiQur&name=pdf)
- Code: [https://hanlab.mit.edu/projects/svdquant](https://hanlab.mit.edu/projects/svdquant)

### Accept Poster

1、Self-supervised contrastive learning performs non-linear system identification

- Paper: [**点击下载**](https://openreview.net/attachment?id=ONfWFluZBI&name=pdf)
- Code: [https://github.com/dynamical-inference/dyncl](https://github.com/dynamical-inference/dyncl)

2、Sparse autoencoders reveal selective remapping of visual concepts during adaptation

- Paper: [**点击下载**](https://openreview.net/attachment?id=imT03YXlG2&name=pdf)
- Code: [https://github.com/dynamical-inference/patchsae](https://github.com/dynamical-inference/patchsae)

3、Multi-Label Test-Time Adaptation with Bound Entropy Minimization

- Paper: [**点击下载**](https://openreview.net/attachment?id=75PhjtbBdr&name=pdf)
- Code: [https://github.com/Jinx630/ML-TTA](https://github.com/Jinx630/ML-TTA)

4、TabM: Advancing tabular deep learning with parameter-efficient ensembling

- Paper: [**点击下载**](https://openreview.net/attachment?id=Sd4wYYOhmY&name=pdf)
- Code: [https://github.com/yandex-research/tabm](https://github.com/yandex-research/tabm)

5、ToolGen: Unified Tool Retrieval and Calling via Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=XLMAMmowdY&name=pdf)
- Code: [https://github.com/Reason-Wang/ToolGen](https://github.com/Reason-Wang/ToolGen)

6、Activation Gradient based Poisoned Sample Detection Against Backdoor Attacks

- Paper: [**点击下载**](https://openreview.net/attachment?id=VNMJfBBUd5&name=pdf)
- Code: [https://github.com/SCLBD/BackdoorBench](https://github.com/SCLBD/BackdoorBench)

7、Video Action Differencing

- Paper: [**点击下载**](https://openreview.net/attachment?id=3bcN6xlO6f&name=pdf)
- Code: [https://huggingface.co/datasets/jmhb/VidDiffBench](https://huggingface.co/datasets/jmhb/VidDiffBench)

8、Optimal Transport for Time Series Imputation

- Paper: [**点击下载**](https://openreview.net/attachment?id=xPTzjpIQNp&name=pdf)
- Code: [https://github.com/FMLYD/PSW-I](https://github.com/FMLYD/PSW-I)

9、RaSA: Rank-Sharing Low-Rank Adaptation

- Paper: [**点击下载**](https://openreview.net/attachment?id=GdXI5zCoAt&name=pdf)
- Code: [https://github.com/zwhe99/RaSA](https://github.com/zwhe99/RaSA)

10、Offline Model-Based Optimization by Learning to Rank

- Paper: [**点击下载**](https://openreview.net/attachment?id=sb1HgVDLjN&name=pdf)
- Code: [https://github.com/lamda-bbo/Offline-RaM](https://github.com/lamda-bbo/Offline-RaM)

11、From Search to Sampling: Generative Models for Robust Algorithmic Recourse

- Paper: [**点击下载**](https://openreview.net/attachment?id=NtwFghsJne&name=pdf)
- Code: [https://github.com/prateekgargx/genre](https://github.com/prateekgargx/genre)

12、Monte Carlo Planning with Large Language Model for Text-Based Game Agents

- Paper: [**点击下载**](https://openreview.net/attachment?id=r1KcapkzCt&name=pdf)
- Code: [https://github.com/winni18/MC-DML](https://github.com/winni18/MC-DML)

13、Robust Root Cause Diagnosis using In-Distribution Interventions

- Paper: [**点击下载**](https://openreview.net/attachment?id=l11DZY5Nxu&name=pdf)
- Code: [https://github.com/nlokeshiisc/IDI_release](https://github.com/nlokeshiisc/IDI_release)

14、Boosting Neural Combinatorial Optimization for Large-Scale Vehicle Routing Problems

- Paper: [**点击下载**](https://openreview.net/attachment?id=TbTJJNjumY&name=pdf)
- Code: [https://github.com/CIAM-Group/SIL](https://github.com/CIAM-Group/SIL)

15、SimulPL: Aligning Human Preferences in Simultaneous Machine Translation

- Paper: [**点击下载**](https://openreview.net/attachment?id=XBF63bHDZw&name=pdf)
- Code: [https://github.com/EurekaForNLP/SimulPL](https://github.com/EurekaForNLP/SimulPL)

16、SCOPE: A Self-supervised Framework for Improving Faithfulness in Conditional Text Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=dTkqaCKLPp&name=pdf)
- Code: [https://github.com/sngdng/scope-faithfulness](https://github.com/sngdng/scope-faithfulness)

17、Clique Number Estimation via Differentiable Functions of Adjacency Matrix Permutations

- Paper: [**点击下载**](https://openreview.net/attachment?id=DFSb67ksVr&name=pdf)
- Code: [https://github.com/structlearning/mxnet](https://github.com/structlearning/mxnet)

18、ADAM: An Embodied Causal Agent in Open-World Environments

- Paper: [**点击下载**](https://openreview.net/attachment?id=Ouu3HnIVBc&name=pdf)
- Code: [https://opencausalab.github.io/ADAM](https://opencausalab.github.io/ADAM)

19、Beware of Calibration Data for Pruning Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=x83w6yGIWb&name=pdf)
- Code: [https://github.com/Dereck0602/calibration_data](https://github.com/Dereck0602/calibration_data)

20、Efficient Residual Learning with Mixture-of-Experts for Universal Dexterous Grasping

- Paper: [**点击下载**](https://openreview.net/attachment?id=BUj9VSCoET&name=pdf)
- Code: [https://sites.google.com/view/resdex](https://sites.google.com/view/resdex)

21、Herald: A Natural Language Annotated Lean 4 Dataset

- Paper: [**点击下载**](https://openreview.net/attachment?id=Se6MgCtRhz&name=pdf)
- Code: [https://huggingface.co/FrenzyMath/Herald_translator](https://huggingface.co/FrenzyMath/Herald_translator)

22、HyperFace: Generating Synthetic Face Recognition Datasets by Exploring Face Embedding Hypersphere

- Paper: [**点击下载**](https://openreview.net/attachment?id=4YzVF9isgD&name=pdf)
- Code: [https://www.idiap.ch/paper/hyperface](https://www.idiap.ch/paper/hyperface)

23、DPLM-2: A Multimodal Diffusion Protein Language Model

- Paper: [**点击下载**](https://openreview.net/attachment?id=5z9GjHgerY&name=pdf)
- Code: [https://bytedance.github.io/dplm/dplm-2](https://bytedance.github.io/dplm/dplm-2)

24、Language Imbalance Driven Rewarding for Multilingual Self-improving

- Paper: [**点击下载**](https://openreview.net/attachment?id=Kak2ZH5Itp&name=pdf)
- Code: [https://github.com/ZNLP/Language-Imbalance-Driven-Rewarding](https://github.com/ZNLP/Language-Imbalance-Driven-Rewarding)

25、Web Agents with World Models: Learning and Leveraging Environment Dynamics in Web Navigation

- Paper: [**点击下载**](https://openreview.net/attachment?id=moWiYJuSGF&name=pdf)
- Code: [https://github.com/kyle8581/WMA-Agents](https://github.com/kyle8581/WMA-Agents)

26、Distribution-Free Data Uncertainty for Neural Network Regression

- Paper: [**点击下载**](https://openreview.net/attachment?id=pDDODPtpx9&name=pdf)
- Code: [https://github.com/proto-n/torch-naut](https://github.com/proto-n/torch-naut)

27、LASeR: Towards Diversified and Generalizable Robot Design with Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=7mlvOHL6qJ&name=pdf)
- Code: [https://github.com/WoodySJR/LASeR](https://github.com/WoodySJR/LASeR)

28、SOO-Bench: Benchmarks for Evaluating the Stability of Offline Black-Box Optimization

- Paper: [**点击下载**](https://openreview.net/attachment?id=bqf0aCF3Dd&name=pdf)
- Code: [https://github.com/zhuyiyi-123/SOO-Bench](https://github.com/zhuyiyi-123/SOO-Bench)

29、Stealthy Shield Defense: A Conditional Mutual Information-Based Approach against Black-Box Model Inversion Attacks

- Paper: [**点击下载**](https://openreview.net/attachment?id=p0DjhjPXl3&name=pdf)
- Code: [https://github.com/ZhuangQu/Stealthy-Shield-Defense](https://github.com/ZhuangQu/Stealthy-Shield-Defense)

30、Be More Diverse than the Most Diverse: Optimal Mixtures of Generative Models via Mixture-UCB Bandit Algorithms

- Paper: [**点击下载**](https://openreview.net/attachment?id=2Chkk5Ye2s&name=pdf)
- Code: [https://github.com/Rezaei-Parham/Mixture-UCB](https://github.com/Rezaei-Parham/Mixture-UCB)

31、Arithmetic Transformers Can Length-Generalize in Both Operand Length and Count

- Paper: [**点击下载**](https://openreview.net/attachment?id=eIgGesYKLG&name=pdf)
- Code: [https://github.com/HanseulJo/position-coupling](https://github.com/HanseulJo/position-coupling)

32、One-for-All Few-Shot Anomaly Detection via Instance-Induced Prompt Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=Zzs3JwknAY&name=pdf)
- Code: [https://github.com/Vanssssry/One-For-All-Few-Shot-Anomaly-Detection](https://github.com/Vanssssry/One-For-All-Few-Shot-Anomaly-Detection)

33、URLOST: Unsupervised Representation Learning without Stationarity or Topology

- Paper: [**点击下载**](https://openreview.net/attachment?id=MBBRHDuiwM&name=pdf)
- Code: [https://github.com/zeyuyun1/URLOST/](https://github.com/zeyuyun1/URLOST/)

34、Charting the Design Space of Neural Graph Representations for Subgraph Matching

- Paper: [**点击下载**](https://openreview.net/attachment?id=5pd78GmXC6&name=pdf)
- Code: [https://github.com/structlearning/neural-subm-design-space ](https://github.com/structlearning/neural-subm-design-space )

35、Distilling Dataset into Neural Field

- Paper: [**点击下载**](https://openreview.net/attachment?id=nCrJD7qPJN&name=pdf)
- Code: [https://github.com/aailab-kaist/DDiF](https://github.com/aailab-kaist/DDiF)

36、SIM: Surface-based fMRI Analysis for Inter-Subject Multimodal Decoding from Movie-Watching Experiments

- Paper: [**点击下载**](https://openreview.net/attachment?id=OJsMGsO6yn&name=pdf)
- Code: [https://github.com/metrics-lab/sim](https://github.com/metrics-lab/sim)

37、SPORTU: A Comprehensive Sports Understanding Benchmark for Multimodal Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=x1yOHtFfDh&name=pdf)
- Code: [https://github.com/chili-lab/SPORTU](https://github.com/chili-lab/SPORTU)

38、Group-robust Sample Reweighting for Subpopulation Shifts via Influence Functions

- Paper: [**点击下载**](https://openreview.net/attachment?id=aQj9Ifxrl6&name=pdf)
- Code: [https://github.com/qiaoruiyt/GSR](https://github.com/qiaoruiyt/GSR)

39、GotenNet: Rethinking Efficient 3D Equivariant Graph Neural Networks

- Paper: [**点击下载**](https://openreview.net/attachment?id=5wxCQDtbMo&name=pdf)
- Code: [https://github.com/sarpaykent/GotenNet](https://github.com/sarpaykent/GotenNet)

40、Efficient Automated Circuit Discovery in Transformers using Contextual Decomposition

- Paper: [**点击下载**](https://openreview.net/attachment?id=41HlN8XYM5&name=pdf)
- Code: [https://github.com/adelaidehsu/CD_Circuit](https://github.com/adelaidehsu/CD_Circuit)

41、How to Evaluate Reward Models for RLHF

- Paper: [**点击下载**](https://openreview.net/attachment?id=cbttLtO94Q&name=pdf)
- Code: [https://github.com/lmarena/PPE](https://github.com/lmarena/PPE)

42、An Efficient Framework for Crediting Data Contributors of Diffusion Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=9EqQC2ct4H&name=pdf)
- Code: [https://github.com/q8888620002/data_attribution](https://github.com/q8888620002/data_attribution)

43、You Only Prune Once: Designing Calibration-Free Model Compression With Policy Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=5RZoYIT3u6&name=pdf)
- Code: [https://github.com/LCS2-IIITD/PruneNet](https://github.com/LCS2-IIITD/PruneNet)

44、SageAttention: Accurate 8-Bit Attention for Plug-and-play Inference Acceleration

- Paper: [**点击下载**](https://openreview.net/attachment?id=OL44KtasKc&name=pdf)
- Code: [https://github.com/thu-ml/SageAttention](https://github.com/thu-ml/SageAttention)

45、FreDF: Learning to Forecast in the Frequency Domain

- Paper: [**点击下载**](https://openreview.net/attachment?id=4A9IdSa1ul&name=pdf)
- Code: [https://github.com/Master-PLC/FreDF](https://github.com/Master-PLC/FreDF)

46、Have the VLMs Lost Confidence? A Study of Sycophancy in VLMs

- Paper: [**点击下载**](https://openreview.net/attachment?id=E2PFv7ad3p&name=pdf)
- Code: [https://github.com/galactic123/Sycophancy_in_VLMs.](https://github.com/galactic123/Sycophancy_in_VLMs.)

47、Generative Adapter: Contextualizing Language Models in Parameters with A Single Forward Pass

- Paper: [**点击下载**](https://openreview.net/attachment?id=bc3sUsS6ck&name=pdf)
- Code: [https://github.com/chentong0/generative-adapter](https://github.com/chentong0/generative-adapter)

48、Ensembling Diffusion Models via Adaptive Feature Aggregation

- Paper: [**点击下载**](https://openreview.net/attachment?id=e32cI4r8Eo&name=pdf)
- Code: [https://github.com/tenvence/afa](https://github.com/tenvence/afa)

49、Model merging with SVD to tie the Knots

- Paper: [**点击下载**](https://openreview.net/attachment?id=67X93aZHII&name=pdf)
- Code: [https://github.com/gstoica27/KnOTS](https://github.com/gstoica27/KnOTS)

50、Making Text Embedders Few-Shot Learners

- Paper: [**点击下载**](https://openreview.net/attachment?id=wfLuiDjQ0u&name=pdf)
- Code: [https://github.com/FlagOpen/FlagEmbedding](https://github.com/FlagOpen/FlagEmbedding)

51、PolyhedronNet: Representation Learning for Polyhedra with Surface-attributed Graph

- Paper: [**点击下载**](https://openreview.net/attachment?id=BpyHIrpUOL&name=pdf)
- Code: [https://github.com/dyu62/3D_polyhedron](https://github.com/dyu62/3D_polyhedron)

52、SafeDiffuser: Safe Planning with Diffusion Probabilistic Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=ig2wk7kK9J&name=pdf)
- Code: [https://safediffuser.github.io/safediffuser/](https://safediffuser.github.io/safediffuser/)

53、Searching for Optimal Solutions with LLMs via Bayesian Optimization

- Paper: [**点击下载**](https://openreview.net/attachment?id=aVfDrl7xDV&name=pdf)
- Code: [https://github.com/amazon-science/BOPRO-ICLR-2025](https://github.com/amazon-science/BOPRO-ICLR-2025)

54、Adversarial Generative Flow Network for Solving Vehicle Routing Problems

- Paper: [**点击下载**](https://openreview.net/attachment?id=tBom4xOW1H&name=pdf)
- Code: [https://github.com/ZHANG-NI/AGFN](https://github.com/ZHANG-NI/AGFN)

55、LongMamba: Enhancing Mamba's Long-Context Capabilities via Training-Free Receptive Field Enlargement

- Paper: [**点击下载**](https://openreview.net/attachment?id=fMbLszVO1H&name=pdf)
- Code: [https://github.com/GATECH-EIC/LongMamba](https://github.com/GATECH-EIC/LongMamba)

56、Lawma: The Power of Specialization for Legal Annotation

- Paper: [**点击下载**](https://openreview.net/attachment?id=7El7K1DoyX&name=pdf)
- Code: [https://github.com/socialfoundations/lawma](https://github.com/socialfoundations/lawma)

57、Stiefel Flow Matching for Moment-Constrained Structure Elucidation

- Paper: [**点击下载**](https://openreview.net/attachment?id=84WmbzikPP&name=pdf)
- Code: [https://github.com/aspuru-guzik-group/stiefelFM](https://github.com/aspuru-guzik-group/stiefelFM)

58、No Preference Left Behind: Group Distributional Preference Optimization

- Paper: [**点击下载**](https://openreview.net/attachment?id=bgpNJBD6Va&name=pdf)
- Code: [https://github.com/BigBinnie/GDPO](https://github.com/BigBinnie/GDPO)

59、Robust Weight Initialization for Tanh Neural Networks with Fixed Point Analysis

- Paper: [**点击下载**](https://openreview.net/attachment?id=Es4RPNDtmq&name=pdf)
- Code: [https://github.com/1HyunwooLee/Tanh-Init](https://github.com/1HyunwooLee/Tanh-Init)

60、Agent S: An Open Agentic Framework that Uses Computers Like a Human

- Paper: [**点击下载**](https://openreview.net/attachment?id=lIVRgt4nLv&name=pdf)
- Code: [https://github.com/simular-ai/Agent-S](https://github.com/simular-ai/Agent-S)

61、Semi-Parametric Retrieval via Binary Bag-of-Tokens Index

- Paper: [**点击下载**](https://openreview.net/attachment?id=l0fn10vSyM&name=pdf)
- Code: [https://github.com/jzhoubu/sidr](https://github.com/jzhoubu/sidr)

62、PN-GAIL: Leveraging Non-optimal Information from Imperfect Demonstrations

- Paper: [**点击下载**](https://openreview.net/attachment?id=0e2pcSxQJS&name=pdf)
- Code: [https://github.com/QiangLiuT/PN-GAIL](https://github.com/QiangLiuT/PN-GAIL)

63、Hyperbolic Genome Embeddings

- Paper: [**点击下载**](https://openreview.net/attachment?id=NkGDNM8LB0&name=pdf)
- Code: [https://github.com/rrkhan/HGE](https://github.com/rrkhan/HGE)

64、RTop-K: Ultra-Fast Row-Wise Top-K Selection for Neural Network Acceleration on GPUs

- Paper: [**点击下载**](https://openreview.net/attachment?id=PHg4rAXFVH&name=pdf)
- Code: [https://anonymous.4open.science/r/RTopK-4DE7](https://anonymous.4open.science/r/RTopK-4DE7)

65、Attributing Culture-Conditioned Generations to Pretraining Corpora

- Paper: [**点击下载**](https://openreview.net/attachment?id=XrsOu4KgDE&name=pdf)
- Code: [https://github.com/huihanlhh/CultureGenAttr](https://github.com/huihanlhh/CultureGenAttr)

66、Unveiling the Secret Recipe: A Guide For Supervised Fine-Tuning Small LLMs

- Paper: [**点击下载**](https://openreview.net/attachment?id=eENHKMTOfW&name=pdf)
- Code: [https://github.com/instructlab/training](https://github.com/instructlab/training)

67、DelTA: An Online Document-Level Translation Agent Based on Multi-Level Memory

- Paper: [**点击下载**](https://openreview.net/attachment?id=hoYFLRNbhc&name=pdf)
- Code: [https://github.com/YutongWang1216/DocMTAgent](https://github.com/YutongWang1216/DocMTAgent)

68、Model Editing as a Robust and Denoised variant of DPO: A Case Study on Toxicity

- Paper: [**点击下载**](https://openreview.net/attachment?id=lOi6FtIwR8&name=pdf)
- Code: [https://github.com/Uppaal/detox-edit](https://github.com/Uppaal/detox-edit)

69、A Common Pitfall of Margin-based Language Model Alignment: Gradient Entanglement

- Paper: [**点击下载**](https://openreview.net/attachment?id=YaBiGjuDiC&name=pdf)
- Code: [https://github.com/HumainLab/Understand_MarginPO](https://github.com/HumainLab/Understand_MarginPO)

70、ManiSkill-HAB: A Benchmark for Low-Level Manipulation in Home Rearrangement Tasks

- Paper: [**点击下载**](https://openreview.net/attachment?id=6bKEWevgSd&name=pdf)
- Code: [http://arth-shukla.github.io/mshab](http://arth-shukla.github.io/mshab)

71、Generalized Behavior Learning from Diverse Demonstrations

- Paper: [**点击下载**](https://openreview.net/attachment?id=Q7EjHroO1w&name=pdf)
- Code: [https://github.com/CORE-Robotics-Lab/GSD](https://github.com/CORE-Robotics-Lab/GSD)

72、DynaMath: A Dynamic Visual Benchmark for Evaluating Mathematical Reasoning Robustness of Vision Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=VOAMTA8jKu&name=pdf)
- Code: [https://dynamath.github.io](https://dynamath.github.io)

73、Neural Stochastic Differential Equations for Uncertainty-Aware Offline RL

- Paper: [**点击下载**](https://openreview.net/attachment?id=hxUMQ4fic3&name=pdf)
- Code: [https://github.com/cevahir-koprulu/sde4mbrltatu/tree/iclr2025](https://github.com/cevahir-koprulu/sde4mbrltatu/tree/iclr2025)

74、Safety-Prioritizing Curricula for Constrained Reinforcement Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=f3QR9TEERH&name=pdf)
- Code: [https://github.com/cevahir-koprulu/safe-curriculum](https://github.com/cevahir-koprulu/safe-curriculum)

75、Towards Federated RLHF with Aggregated Client Preference for LLMs

- Paper: [**点击下载**](https://openreview.net/attachment?id=mqNKiEB6pd&name=pdf)
- Code: [https://github.com/HarliWu/FedBiscuit](https://github.com/HarliWu/FedBiscuit)

76、MetaDesigner: Advancing Artistic Typography through AI-Driven, User-Centric, and Multilingual WordArt Synthesis

- Paper: [**点击下载**](https://openreview.net/attachment?id=Mv3GAYJGcW&name=pdf)
- Code: [https://modelscope.cn/studios/WordArt/WordArt](https://modelscope.cn/studios/WordArt/WordArt)

77、Physics of Language Models: Part 3.2, Knowledge Manipulation

- Paper: [**点击下载**](https://openreview.net/attachment?id=oDbiL9CLoS&name=pdf)
- Code: [https://youtu.be/yBL7J0kgldU](https://youtu.be/yBL7J0kgldU)

78、On Calibration of LLM-based Guard Models for Reliable Content Moderation

- Paper: [**点击下载**](https://openreview.net/attachment?id=wUbum0nd9N&name=pdf)
- Code: [https://github.com/Waffle-Liu/calibration_guard_model](https://github.com/Waffle-Liu/calibration_guard_model)

79、Physics of Language Models: Part 2.2, How to Learn From Mistakes on Grade-School Math Problems

- Paper: [**点击下载**](https://openreview.net/attachment?id=zpDGwcmMV4&name=pdf)
- Code: [https://youtu.be/yBL7J0kgldU](https://youtu.be/yBL7J0kgldU)

80、Causal Order: The Key to Leveraging Imperfect Experts in Causal Inference

- Paper: [**点击下载**](https://openreview.net/attachment?id=9juyeCqL0u&name=pdf)
- Code: [https://github.com/AniketVashishtha/Causal_Order_Imperfect_Experts](https://github.com/AniketVashishtha/Causal_Order_Imperfect_Experts)

81、Physics of Language Models: Part 2.1, Grade-School Math and the Hidden Reasoning Process

- Paper: [**点击下载**](https://openreview.net/attachment?id=Tn5B6Udq3E&name=pdf)
- Code: [https://youtu.be/yBL7J0kgldU](https://youtu.be/yBL7J0kgldU)

82、Agent-Oriented Planning in Multi-Agent Systems

- Paper: [**点击下载**](https://openreview.net/attachment?id=EqcLAU6gyU&name=pdf)
- Code: [https://github.com/lalaliat/Agent-Oriented-Planning](https://github.com/lalaliat/Agent-Oriented-Planning)

83、Diffusion Generative Modeling for Spatially Resolved Gene Expression Inference from Histology Images

- Paper: [**点击下载**](https://openreview.net/attachment?id=FtjLUHyZAO&name=pdf)
- Code: [https://github.com/SichenZhu/Stem](https://github.com/SichenZhu/Stem)

84、Attention in Large Language Models Yields Efficient Zero-Shot Re-Rankers

- Paper: [**点击下载**](https://openreview.net/attachment?id=yzloNYH3QN&name=pdf)
- Code: [https://github.com/OSU-NLP-Group/In-Context-Reranking](https://github.com/OSU-NLP-Group/In-Context-Reranking)

85、ExACT: Teaching AI Agents to Explore with Reflective-MCTS and Exploratory Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=GBIUbwW9D8&name=pdf)
- Code: [https://aka.ms/ExACT](https://aka.ms/ExACT)

86、STAFF: Speculative Coreset Selection for Task-Specific Fine-tuning

- Paper: [**点击下载**](https://openreview.net/attachment?id=FAfxvdv1Dy&name=pdf)
- Code: [https://github.com/shiningrain/STAFF](https://github.com/shiningrain/STAFF)

87、Context Clues: Evaluating Long Context Models for Clinical Prediction Tasks on EHR Data

- Paper: [**点击下载**](https://openreview.net/attachment?id=zg3ec1TdAP&name=pdf)
- Code: [https://github.com/som-shahlab/long_context_clues](https://github.com/som-shahlab/long_context_clues)

88、Scale-Aware Contrastive Reverse Distillation for Unsupervised Medical Anomaly Detection

- Paper: [**点击下载**](https://openreview.net/attachment?id=HNOo4UNPBF&name=pdf)
- Code: [https://github.com/MedAITech/SCRD4AD](https://github.com/MedAITech/SCRD4AD)

89、Learning Efficient Positional Encodings with Graph Neural Networks

- Paper: [**点击下载**](https://openreview.net/attachment?id=AWg2tkbydO&name=pdf)
- Code: [https://github.com/ehejin/Pearl-PE](https://github.com/ehejin/Pearl-PE)

90、Scalable Influence and Fact Tracing for Large Language Model Pretraining

- Paper: [**点击下载**](https://openreview.net/attachment?id=gLa96FlWwn&name=pdf)
- Code: [https://github.com/pair-code/pretraining-tda](https://github.com/pair-code/pretraining-tda)

91、Mastering Task Arithmetic: $\tau$Jp as a Key Indicator for Weight Disentanglement

- Paper: [**点击下载**](https://openreview.net/attachment?id=1VwWi6zbxs&name=pdf)
- Code: [https://github.com/katoro8989/tau-Jp_Task_Arithmetic](https://github.com/katoro8989/tau-Jp_Task_Arithmetic)

92、Diffusion State-Guided Projected Gradient for Inverse Problems

- Paper: [**点击下载**](https://openreview.net/attachment?id=kRBQwlkFSP&name=pdf)
- Code: [https://github.com/Anima-Lab/DiffStateGrad](https://github.com/Anima-Lab/DiffStateGrad)

93、PALMBENCH: A COMPREHENSIVE BENCHMARK OF COMPRESSED LARGE LANGUAGE MODELS ON MOBILE PLATFORMS

- Paper: [**点击下载**](https://openreview.net/attachment?id=xzSUdw6s76&name=pdf)
- Code: [https://github.com/JimmyLi-Network/PalmBench.github.io/](https://github.com/JimmyLi-Network/PalmBench.github.io/)

94、Generating CAD Code with Vision-Language Models for 3D Designs

- Paper: [**点击下载**](https://openreview.net/attachment?id=BLWaTeucYX&name=pdf)
- Code: [https://github.com/Kamel773/CAD_Code_Generation](https://github.com/Kamel773/CAD_Code_Generation)

95、MaxInfoRL: Boosting exploration in reinforcement learning through information gain maximization

- Paper: [**点击下载**](https://openreview.net/attachment?id=R4q3cY3kQf&name=pdf)
- Code: [https://sukhijab.github.io/projects/maxinforl/](https://sukhijab.github.io/projects/maxinforl/)

96、Forgetting Transformer: Softmax Attention with a Forget Gate

- Paper: [**点击下载**](https://openreview.net/attachment?id=q2Lnyegkr8&name=pdf)
- Code: [https://github.com/zhixuan-lin/forgetting-transformer](https://github.com/zhixuan-lin/forgetting-transformer)

97、ReMoE: Fully Differentiable Mixture-of-Experts with ReLU Routing

- Paper: [**点击下载**](https://openreview.net/attachment?id=4D0f16Vwc3&name=pdf)
- Code: [https://github.com/thu-ml/ReMoE](https://github.com/thu-ml/ReMoE)

98、GSM-Symbolic: Understanding the Limitations of Mathematical Reasoning in Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=AjXkRZIvjB&name=pdf)
- Code: [https://github.com/apple/ml-gsm-symbolic](https://github.com/apple/ml-gsm-symbolic)

99、Score-based Self-supervised MRI Denoising

- Paper: [**点击下载**](https://openreview.net/attachment?id=uNd289HjLi&name=pdf)
- Code: [https://jiachentu.github.io/Corruption2Self-Self-Supervised-Denoising/](https://jiachentu.github.io/Corruption2Self-Self-Supervised-Denoising/)

100、ProteinBench: A Holistic Evaluation of Protein Foundation Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=BksqWM8737&name=pdf)
- Code: [https://proteinbench.github.io/](https://proteinbench.github.io/)

101、LeanQuant: Accurate and Scalable Large Language Model Quantization with Loss-error-aware Grid

- Paper: [**点击下载**](https://openreview.net/attachment?id=ISqx8giekS&name=pdf)
- Code: [https://github.com/LeanModels/LeanQuant](https://github.com/LeanModels/LeanQuant)

102、Strategist: Self-improvement of LLM Decision Making via Bi-Level Tree Search

- Paper: [**点击下载**](https://openreview.net/attachment?id=gfI9v7AbFg&name=pdf)
- Code: [https://llm-strategist.github.io](https://llm-strategist.github.io)

103、GReaTer: Gradients Over Reasoning Makes Smaller Language Models Strong Prompt Optimizers

- Paper: [**点击下载**](https://openreview.net/attachment?id=fWRBheSJth&name=pdf)
- Code: [https://github.com/psunlpgroup/GreaTer](https://github.com/psunlpgroup/GreaTer)

104、LANTERN: Accelerating Visual Autoregressive Models with Relaxed Speculative Decoding

- Paper: [**点击下载**](https://openreview.net/attachment?id=98d7DLMGdt&name=pdf)
- Code: [https://github.com/jadohu/LANTERN](https://github.com/jadohu/LANTERN)

105、Follow My Instruction and Spill the Beans: Scalable Data Extraction from Retrieval-Augmented Generation Systems

- Paper: [**点击下载**](https://openreview.net/attachment?id=Y4aWwRh25b&name=pdf)
- Code: [https://github.com/zhentingqi/rag-privacy.git](https://github.com/zhentingqi/rag-privacy.git)

106、EditRoom: LLM-parameterized Graph Diffusion for Composable 3D Room Layout Editing

- Paper: [**点击下载**](https://openreview.net/attachment?id=Y2Dh8rWwlb&name=pdf)
- Code: [https://eric-ai-lab.github.io/edit-room.github.io/](https://eric-ai-lab.github.io/edit-room.github.io/)

107、MrT5: Dynamic Token Merging for Efficient Byte-level Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=VYWBMq1L7H&name=pdf)
- Code: [https://github.com/jkallini/mrt5](https://github.com/jkallini/mrt5)

108、Simple is Effective: The Roles of Graphs and Large Language Models in Knowledge-Graph-Based Retrieval-Augmented Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=JvkuZZ04O7&name=pdf)
- Code: [https://github.com/Graph-COM/SubgraphRAG](https://github.com/Graph-COM/SubgraphRAG)

109、What's the Move? Hybrid Imitation Learning via Salient Points

- Paper: [**点击下载**](https://openreview.net/attachment?id=r0pLGGcuY6&name=pdf)
- Code: [http://sphinx-manip.github.io](http://sphinx-manip.github.io)

110、LocoVR: Multiuser Indoor Locomotion Dataset in Virtual Reality

- Paper: [**点击下载**](https://openreview.net/attachment?id=9mBodivRIo&name=pdf)
- Code: [https://github.com/kt2024-hal/LocoVR](https://github.com/kt2024-hal/LocoVR)

111、Greener GRASS: Enhancing GNNs with Encoding, Rewiring, and Attention

- Paper: [**点击下载**](https://openreview.net/attachment?id=rEQqBZIz49&name=pdf)
- Code: [https://orcid.org/0009-0003-5683-4645](https://orcid.org/0009-0003-5683-4645)

112、Conformal Language Model Reasoning with Coherent Factuality

- Paper: [**点击下载**](https://openreview.net/attachment?id=AJpUZd8Clb&name=pdf)
- Code: [https://github.com/maxrubintoles/Conformal_LM_Reasoning](https://github.com/maxrubintoles/Conformal_LM_Reasoning)

113、Dissecting Adversarial Robustness of Multimodal LM Agents

- Paper: [**点击下载**](https://openreview.net/attachment?id=YauQYh2k1g&name=pdf)
- Code: [https://github.com/ChenWu98/agent-attack](https://github.com/ChenWu98/agent-attack)

114、ACC-Collab: An Actor-Critic Approach to Multi-Agent LLM Collaboration

- Paper: [**点击下载**](https://openreview.net/attachment?id=nfKfAzkiez&name=pdf)
- Code: [https://github.com/LlenRotse/ACC-Collab](https://github.com/LlenRotse/ACC-Collab)

115、More Experts Than Galaxies: Conditionally-Overlapping Experts with Biologically-Inspired Fixed Routing

- Paper: [**点击下载**](https://openreview.net/attachment?id=1qq1QJKM5q&name=pdf)
- Code: [https://github.com/Shaier/COMET.git](https://github.com/Shaier/COMET.git)

116、In-context Time Series Predictor

- Paper: [**点击下载**](https://openreview.net/attachment?id=dCcY2pyNIO&name=pdf)
- Code: [https://anonymous.4open.science/r/ICTSP-C995](https://anonymous.4open.science/r/ICTSP-C995)

117、Discovering Influential Neuron Path in Vision Transformers

- Paper: [**点击下载**](https://openreview.net/attachment?id=WQQyJbr5Lh&name=pdf)
- Code: [https://foundation-model-research.github.io/NeuronPath/](https://foundation-model-research.github.io/NeuronPath/)

118、Directional Gradient Projection for Robust Fine-Tuning of Foundation Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=goBaGHLAdP&name=pdf)
- Code: [https://github.com/chengyuehuang511/DiGraP](https://github.com/chengyuehuang511/DiGraP)

119、AgentOccam: A Simple Yet Strong Baseline for LLM-Based Web Agents

- Paper: [**点击下载**](https://openreview.net/attachment?id=oWdzUpOlkX&name=pdf)
- Code: [https://github.com/amazon-science/AgentOccam](https://github.com/amazon-science/AgentOccam)

120、Probe Pruning: Accelerating LLMs through Dynamic Pruning via Model-Probing

- Paper: [**点击下载**](https://openreview.net/attachment?id=WOt1owGfuN&name=pdf)
- Code: [https://github.com/Qi-Le1/Probe_Pruning](https://github.com/Qi-Le1/Probe_Pruning)

121、L3Ms — Lagrange Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=ULGbw2URE3&name=pdf)
- Code: [https://www.reddit.com/r/LocalLLaMA/comments/180p17f/new_claude_21_refuses_to_kill_a_python_process/](https://www.reddit.com/r/LocalLLaMA/comments/180p17f/new_claude_21_refuses_to_kill_a_python_process/)

122、ClimaQA: An Automated Evaluation Framework for Climate Question Answering Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=goFpCuJalN&name=pdf)
- Code: [https://github.com/Rose-STL-Lab/genie-climaqa](https://github.com/Rose-STL-Lab/genie-climaqa)

123、Human-inspired Episodic Memory for Infinite Context LLMs

- Paper: [**点击下载**](https://openreview.net/attachment?id=BI2int5SAC&name=pdf)
- Code: [https://github.com/em-llm/EM-LLM-model](https://github.com/em-llm/EM-LLM-model)

124、ORSO: Accelerating Reward Design via Online Reward Selection and Policy Optimization

- Paper: [**点击下载**](https://openreview.net/attachment?id=0uRc3CfJIQ&name=pdf)
- Code: [https://github.com/Improbable-AI/orso](https://github.com/Improbable-AI/orso)

125、Object-Centric Pretraining via Target Encoder Bootstrapping

- Paper: [**点击下载**](https://openreview.net/attachment?id=7d2JwGbxhA&name=pdf)
- Code: [https://github.com/djukicn/ocebo](https://github.com/djukicn/ocebo)

126、Lossy Compression with Pretrained Diffusion Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=raUnLe0Z04&name=pdf)
- Code: [https://github.com/jeremyiv/diffc](https://github.com/jeremyiv/diffc)

127、Modality-Specialized Synergizers for Interleaved Vision-Language Generalists

- Paper: [**点击下载**](https://openreview.net/attachment?id=7UgQjFEadn&name=pdf)
- Code: [https://github.com/VT-NLP/MoSS](https://github.com/VT-NLP/MoSS)

128、Self-Normalized Resets for Plasticity in Continual Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=G82uQztzxl&name=pdf)
- Code: [https://github.com/ajozefiak/SelfNormalizedResets](https://github.com/ajozefiak/SelfNormalizedResets)

129、Rapidly Adapting Policies to the Real-World via Simulation-Guided Fine-Tuning

- Paper: [**点击下载**](https://openreview.net/attachment?id=XwUrzurG94&name=pdf)
- Code: [https://weirdlabuw.github.io/sgft/](https://weirdlabuw.github.io/sgft/)

130、CoRNStack: High-Quality Contrastive Data for Better Code Retrieval and Reranking

- Paper: [**点击下载**](https://openreview.net/attachment?id=iyJOUELYir&name=pdf)
- Code: [https://github.com/gangiswag/cornstack](https://github.com/gangiswag/cornstack)

131、On Linear Representations and Pretraining Data Frequency in Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=EDoD3DgivF&name=pdf)
- Code: [https://github.com/allenai/freq](https://github.com/allenai/freq)

132、Quantifying Generalization Complexity for Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=jpSLXoRKnH&name=pdf)
- Code: [https://github.com/zhentingqi/scylla](https://github.com/zhentingqi/scylla)

133、Real2Code: Reconstruct Articulated Objects via Code Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=CAssIgPN4I&name=pdf)
- Code: [https://real2code.github.io/](https://real2code.github.io/)

134、{$\tau$}-bench: A Benchmark for \underline{T}ool-\underline{A}gent-\underline{U}ser Interaction in Real-World Domains

- Paper: [**点击下载**](https://openreview.net/attachment?id=roNSXZpUDN&name=pdf)
- Code: [https://github.com/sierra-research/tau-bench](https://github.com/sierra-research/tau-bench)

135、DEPfold: RNA Secondary Structure Prediction as Dependency Parsing.

- Paper: [**点击下载**](https://openreview.net/attachment?id=DpLFmc09pC&name=pdf)
- Code: [https://github.com/Vicky-0256/DEPfold.git](https://github.com/Vicky-0256/DEPfold.git)

136、What Matters in Learning from Large-Scale Datasets for Robot Manipulation

- Paper: [**点击下载**](https://openreview.net/attachment?id=LqhorpRLIm&name=pdf)
- Code: [https://robo-mimiclabs.github.io/](https://robo-mimiclabs.github.io/)

137、APE: Faster and Longer Context-Augmented Generation via Adaptive Parallel Encoding

- Paper: [**点击下载**](https://openreview.net/attachment?id=yUC8pU508S&name=pdf)
- Code: [https://github.com/Infini-AI-Lab/APE](https://github.com/Infini-AI-Lab/APE)

138、Differentiable Optimization of Similarity Scores Between Models and Brains

- Paper: [**点击下载**](https://openreview.net/attachment?id=vWRwdmA3wU&name=pdf)
- Code: [https://nacloos.github.io/diffscore-site](https://nacloos.github.io/diffscore-site)

139、Chunk-Distilled Language Modeling

- Paper: [**点击下载**](https://openreview.net/attachment?id=nrvoWOWcyg&name=pdf)
- Code: [https://github.com/yanhong-lbh/cd-lm](https://github.com/yanhong-lbh/cd-lm)

140、GOttack: Universal Adversarial Attacks on Graph Neural Networks via Graph Orbits Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=YbURbViE7l&name=pdf)
- Code: [https://github.com/cakcora/GOttack](https://github.com/cakcora/GOttack)

141、NutriBench: A Dataset for Evaluating Large Language Models in Nutrition Estimation from Meal Descriptions

- Paper: [**点击下载**](https://openreview.net/attachment?id=6LtdZCyuZR&name=pdf)
- Code: [https://mehak126.github.io/nutribench.html](https://mehak126.github.io/nutribench.html)

142、The Value of Sensory Information to a Robot

- Paper: [**点击下载**](https://openreview.net/attachment?id=ikr5XomWHS&name=pdf)
- Code: [https://sites.google.com/view/vosi-robotics/](https://sites.google.com/view/vosi-robotics/)

143、Explore Theory of Mind: program-guided adversarial data generation for theory of mind reasoning

- Paper: [**点击下载**](https://openreview.net/attachment?id=246rHKUnnf&name=pdf)
- Code: [https://github.com/facebookresearch/exploretom](https://github.com/facebookresearch/exploretom)

144、HelpSteer2-Preference: Complementing Ratings with Preferences

- Paper: [**点击下载**](https://openreview.net/attachment?id=MnfHxPP5gs&name=pdf)
- Code: [https://huggingface.co/datasets/nvidia/HelpSteer2#preferences-new---1-oct-2024](https://huggingface.co/datasets/nvidia/HelpSteer2#preferences-new---1-oct-2024)

145、Optimizing 4D Gaussians for Dynamic Scene Video from Single Landscape Images

- Paper: [**点击下载**](https://openreview.net/attachment?id=IcYDRzcccP&name=pdf)
- Code: [https://cvsp-lab.github.io/ICLR2025_3D-MOM/](https://cvsp-lab.github.io/ICLR2025_3D-MOM/)

146、Samba: Simple Hybrid State Space Models for Efficient Unlimited Context Language Modeling

- Paper: [**点击下载**](https://openreview.net/attachment?id=bIlnpVM4bc&name=pdf)
- Code: [https://github.com/microsoft/Samba](https://github.com/microsoft/Samba)

147、Palu: KV-Cache Compression with Low-Rank Projection

- Paper: [**点击下载**](https://openreview.net/attachment?id=LWMS4pk2vK&name=pdf)
- Code: [https://github.com/shadowpa0327/Palu](https://github.com/shadowpa0327/Palu)

148、Shared-AE: Automatic Identification of Shared Subspaces in High-dimensional Neural and Behavioral Activity

- Paper: [**点击下载**](https://openreview.net/attachment?id=zXCnIyX9MG&name=pdf)
- Code: [https://github.com/saxenalab-neuro/Shared-AE](https://github.com/saxenalab-neuro/Shared-AE)

149、Many-Objective Multi-Solution Transport

- Paper: [**点击下载**](https://openreview.net/attachment?id=Neb17mimVH&name=pdf)
- Code: [https://github.com/tianyi-lab/MosT](https://github.com/tianyi-lab/MosT)

150、Latent Safety-Constrained Policy Approach for Safe Offline Reinforcement Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=bDt5qc7TfO&name=pdf)
- Code: [https://github.com/PrajwalKoirala/LSPC-Safe-Offline-RL](https://github.com/PrajwalKoirala/LSPC-Safe-Offline-RL)

151、Learning to Explore and Exploit with GNNs for Unsupervised Combinatorial Optimization

- Paper: [**点击下载**](https://openreview.net/attachment?id=vaJ4FObpXN&name=pdf)
- Code: [https://github.com/utkuumur/X2GNN](https://github.com/utkuumur/X2GNN)

152、InterMask: 3D Human Interaction Generation via Collaborative Masked Modeling

- Paper: [**点击下载**](https://openreview.net/attachment?id=ZAyuwJYN8N&name=pdf)
- Code: [https://gohar-malik.github.io/intermask/](https://gohar-malik.github.io/intermask/)

153、Can Knowledge Editing Really Correct Hallucinations?

- Paper: [**点击下载**](https://openreview.net/attachment?id=hmDt068MoZ&name=pdf)
- Code: [https://llm-editing.github.io](https://llm-editing.github.io)

154、BoneMet: An Open Large-Scale Multi-Modal Murine Dataset for Breast Cancer Bone Metastasis Diagnosis and Prognosis

- Paper: [**点击下载**](https://openreview.net/attachment?id=YH4M1Tbxfz&name=pdf)
- Code: [https://huggingface.co/datasets/BoneMet/BoneMet](https://huggingface.co/datasets/BoneMet/BoneMet)

155、ALLaM: Large Language Models for Arabic and English

- Paper: [**点击下载**](https://openreview.net/attachment?id=MscdsFVZrN&name=pdf)
- Code: [https://huggingface.co/ALLaM-AI](https://huggingface.co/ALLaM-AI)

156、Improving Graph Neural Networks by Learning Continuous Edge Directions

- Paper: [**点击下载**](https://openreview.net/attachment?id=iAmR7FfMmq&name=pdf)
- Code: [https://github.com/hormoz-lab/coed-gnn](https://github.com/hormoz-lab/coed-gnn)

157、SiReRAG: Indexing Similar and Related Information for Multihop Reasoning

- Paper: [**点击下载**](https://openreview.net/attachment?id=yp95goUAT1&name=pdf)
- Code: [https://github.com/SalesforceAIResearch/SiReRAG](https://github.com/SalesforceAIResearch/SiReRAG)

158、Eliciting Human Preferences with Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=LvDwwAgMEW&name=pdf)
- Code: [https://github.com/alextamkin/generative-elicitation](https://github.com/alextamkin/generative-elicitation)

159、Preserving Deep Representations in One-Shot Pruning: A Hessian-Free Second-Order Optimization Framework

- Paper: [**点击下载**](https://openreview.net/attachment?id=eNQp79A5Oz&name=pdf)
- Code: [https://github.com/mazumder-lab/SNOWS](https://github.com/mazumder-lab/SNOWS)

160、Chemistry-Inspired Diffusion with Non-Differentiable Guidance

- Paper: [**点击下载**](https://openreview.net/attachment?id=4dAgG8ma3B&name=pdf)
- Code: [https://github.com/A-Chicharito-S/ChemGuide](https://github.com/A-Chicharito-S/ChemGuide)

161、Aligning Language Models with Demonstrated Feedback

- Paper: [**点击下载**](https://openreview.net/attachment?id=1qGkuxI9UX&name=pdf)
- Code: [https://github.com/SALT-NLP/demonstrated-feedback](https://github.com/SALT-NLP/demonstrated-feedback)

162、Efficient Diffusion Transformer Policies with Mixture of Expert Denoisers for Multitask Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=nDmwloEl3N&name=pdf)
- Code: [https://mbreuss.github.io/MoDE_Diffusion_Policy/](https://mbreuss.github.io/MoDE_Diffusion_Policy/)

163、SymmCD: Symmetry-Preserving Crystal Generation with Diffusion Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=xnssGv9rpW&name=pdf)
- Code: [https://github.com/sibasmarak/SymmCD/](https://github.com/sibasmarak/SymmCD/)

164、InvestESG: A multi-agent reinforcement learning benchmark for studying climate investment as a social dilemma

- Paper: [**点击下载**](https://openreview.net/attachment?id=2TasVD7FXp&name=pdf)
- Code: [https://github.com/yuanjiayiy/InvestESG](https://github.com/yuanjiayiy/InvestESG)

165、The Pitfalls of Memorization: When Memorization Hurts Generalization

- Paper: [**点击下载**](https://openreview.net/attachment?id=vVhZh9ZpIM&name=pdf)
- Code: [https://github.com/facebookresearch/Pitfalls-of-Memorization](https://github.com/facebookresearch/Pitfalls-of-Memorization)

166、See It from My Perspective: How Language Affects Cultural Bias in Image Understanding

- Paper: [**点击下载**](https://openreview.net/attachment?id=Xbl6t6zxZs&name=pdf)
- Code: [https://github.com/amith-ananthram/see-it-from-my-perspective](https://github.com/amith-ananthram/see-it-from-my-perspective)

167、Cocoon: Robust Multi-Modal Perception with Uncertainty-Aware Sensor Fusion

- Paper: [**点击下载**](https://openreview.net/attachment?id=DKgAFfCs5F&name=pdf)
- Code: [https://minkyoungcho.github.io/cocoon/](https://minkyoungcho.github.io/cocoon/)

168、Benchmarking Vision Language Model Unlearning via Fictitious Facial Identity Dataset

- Paper: [**点击下载**](https://openreview.net/attachment?id=0y3hGn1wOk&name=pdf)
- Code: [https://huggingface.co/datasets/gray311/FIUBench](https://huggingface.co/datasets/gray311/FIUBench)

169、Discrete GCBF Proximal Policy Optimization for Multi-agent Safe Optimal Control

- Paper: [**点击下载**](https://openreview.net/attachment?id=1X1R7P6yzt&name=pdf)
- Code: [https://mit-realm.github.io/dgppo/](https://mit-realm.github.io/dgppo/)

170、Cauchy-Schwarz Regularizers

- Paper: [**点击下载**](https://openreview.net/attachment?id=KZu3xhPhke&name=pdf)
- Code: [https://github.com/IIP-Group/CS_regularizers](https://github.com/IIP-Group/CS_regularizers)

171、HELMET: How to Evaluate Long-context Models Effectively and Thoroughly

- Paper: [**点击下载**](https://openreview.net/attachment?id=293V3bJbmE&name=pdf)
- Code: [https://princeton-nlp.github.io/HELMET](https://princeton-nlp.github.io/HELMET)

172、R-Sparse: Rank-Aware Activation Sparsity for Efficient LLM Inference

- Paper: [**点击下载**](https://openreview.net/attachment?id=9VMW4iXfKt&name=pdf)
- Code: [https://github.com/VITA-Group/R-Sparse](https://github.com/VITA-Group/R-Sparse)

173、A Unified Framework for Forward and Inverse Problems in Subsurface Imaging using Latent Space Translations

- Paper: [**点击下载**](https://openreview.net/attachment?id=yIlyHJdYV3&name=pdf)
- Code: [https://github.com/KGML-lab/Generalized-Forward-Inverse-Framework-for-DL4SI](https://github.com/KGML-lab/Generalized-Forward-Inverse-Framework-for-DL4SI)

174、Regressing the Relative Future: Efficient Policy Optimization for Multi-turn RLHF

- Paper: [**点击下载**](https://openreview.net/attachment?id=cVyELMpMRS&name=pdf)
- Code: [https://github.com/ZhaolinGao/REFUEL/](https://github.com/ZhaolinGao/REFUEL/)

175、MM-EMBED: UNIVERSAL MULTIMODAL RETRIEVAL WITH MULTIMODAL LLMS

- Paper: [**点击下载**](https://openreview.net/attachment?id=i45NQb2iKO&name=pdf)
- Code: [https://huggingface.co/nvidia/MM-Embed](https://huggingface.co/nvidia/MM-Embed)

176、Procedural Synthesis of Synthesizable Molecules

- Paper: [**点击下载**](https://openreview.net/attachment?id=OGfyzExd69&name=pdf)
- Code: [https://github.com/shiningsunnyday/SynthesisNet](https://github.com/shiningsunnyday/SynthesisNet)

177、Instruct-SkillMix: A Powerful Pipeline for LLM Instruction Tuning

- Paper: [**点击下载**](https://openreview.net/attachment?id=44z7HL4mfX&name=pdf)
- Code: [https://github.com/princeton-pli/Instruct-SkillMix](https://github.com/princeton-pli/Instruct-SkillMix)

178、Adapters for Altering LLM Vocabularies: What Languages Benefit the Most?

- Paper: [**点击下载**](https://openreview.net/attachment?id=KxQRHOre9D&name=pdf)
- Code: [https://github.com/h-j-han/VocADT](https://github.com/h-j-han/VocADT)

179、DOTS: Learning to Reason Dynamically in LLMs via Optimal Reasoning Trajectories Search

- Paper: [**点击下载**](https://openreview.net/attachment?id=tn2mjzjSyR&name=pdf)
- Code: [https://github.com/MurongYue/DOTS](https://github.com/MurongYue/DOTS)

180、Teaching LLMs How to Learn with Contextual Fine-Tuning

- Paper: [**点击下载**](https://openreview.net/attachment?id=FS2nukC2jv&name=pdf)
- Code: [https://younwoochoi.github.io/cft-iclr/](https://younwoochoi.github.io/cft-iclr/)

181、An Undetectable Watermark for Generative Image Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=jlhBFm7T2J&name=pdf)
- Code: [https://github.com/XuandongZhao/PRC-Watermark](https://github.com/XuandongZhao/PRC-Watermark)

182、STRAP: Robot Sub-Trajectory Retrieval for Augmented Policy Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=4VHiptx7xe&name=pdf)
- Code: [https://strapaper.github.io/strap.github.io/](https://strapaper.github.io/strap.github.io/)

183、Residual Stream Analysis with Multi-Layer SAEs

- Paper: [**点击下载**](https://openreview.net/attachment?id=XAjfjizaKs&name=pdf)
- Code: [https://github.com/tim-lawson/mlsae](https://github.com/tim-lawson/mlsae)

184、Trivialized Momentum Facilitates Diffusion Generative Modeling on Lie Groups

- Paper: [**点击下载**](https://openreview.net/attachment?id=DTatjJTDl1&name=pdf)
- Code: [https://github.com/yuchen-zhu-zyc/TDM](https://github.com/yuchen-zhu-zyc/TDM)

185、Towards Fast, Specialized Machine Learning Force Fields: Distilling Foundation Models via Energy Hessians

- Paper: [**点击下载**](https://openreview.net/attachment?id=1durmugh3I&name=pdf)
- Code: [https://github.com/ASK-Berkeley/MLFF-distill](https://github.com/ASK-Berkeley/MLFF-distill)

186、On the Transfer of Object-Centric Representation Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=bSq0XGS3kW&name=pdf)
- Code: [https://rw-ocrl.github.io/ftdinosaur-paper/](https://rw-ocrl.github.io/ftdinosaur-paper/)

187、Few-Class Arena: A Benchmark for Efficient Selection of Vision Models and Dataset Difficulty Measurement

- Paper: [**点击下载**](https://openreview.net/attachment?id=2ET561DyPe&name=pdf)
- Code: [https://github.com/bryanbocao/fca](https://github.com/bryanbocao/fca)

188、OvercookedV2: Rethinking Overcooked for Zero-Shot Coordination

- Paper: [**点击下载**](https://openreview.net/attachment?id=hlvLM3GX8R&name=pdf)
- Code: [https://github.com/FLAIROx/JaxMARL](https://github.com/FLAIROx/JaxMARL)

189、Dataset Distillation via Knowledge Distillation: Towards Efficient Self-Supervised Pre-training of Deep Networks

- Paper: [**点击下载**](https://openreview.net/attachment?id=c61unr33XA&name=pdf)
- Code: [https://github.com/BigML-CS-UCLA/MKDT](https://github.com/BigML-CS-UCLA/MKDT)

190、MA-RLHF: Reinforcement Learning from Human Feedback with Macro Actions

- Paper: [**点击下载**](https://openreview.net/attachment?id=WWXjMYZxfH&name=pdf)
- Code: [https://github.com/ernie-research/MA-RLHF](https://github.com/ernie-research/MA-RLHF)

191、Efficient Active Imitation Learning with Random Network Distillation

- Paper: [**点击下载**](https://openreview.net/attachment?id=GFgn2LprFR&name=pdf)
- Code: [https://sites.google.com/view/rnd-dagger](https://sites.google.com/view/rnd-dagger)

192、Discrete Diffusion Schrödinger Bridge Matching for Graph Transformation

- Paper: [**点击下载**](https://openreview.net/attachment?id=tQyh0gnfqW&name=pdf)
- Code: [https://github.com/junhkim1226/DDSBM](https://github.com/junhkim1226/DDSBM)

193、MatExpert: Decomposing Materials Discovery By Mimicking Human Experts

- Paper: [**点击下载**](https://openreview.net/attachment?id=AUBvo4sxVL&name=pdf)
- Code: [https://github.com/BangLab-UdeM-Mila/MatExpert](https://github.com/BangLab-UdeM-Mila/MatExpert)

194、Fugatto 1: Foundational Generative Audio Transformer Opus 1

- Paper: [**点击下载**](https://openreview.net/attachment?id=B2Fqu7Y2cd&name=pdf)
- Code: [https://fugatto.github.io/](https://fugatto.github.io/)

195、Consistency Models Made Easy

- Paper: [**点击下载**](https://openreview.net/attachment?id=xQVxo9dSID&name=pdf)
- Code: [https://github.com/locuslab/ect](https://github.com/locuslab/ect)

196、AttriBoT: A Bag of Tricks for Efficiently Approximating Leave-One-Out Context Attribution

- Paper: [**点击下载**](https://openreview.net/attachment?id=9kJperA2a4&name=pdf)
- Code: [https://github.com/r-three/AttriBoT](https://github.com/r-three/AttriBoT)

197、ElasticTok: Adaptive Tokenization for Image and Video

- Paper: [**点击下载**](https://openreview.net/attachment?id=tFV5GrWOGm&name=pdf)
- Code: [http://largeworldmodel.github.io/elastictok](http://largeworldmodel.github.io/elastictok)

198、Beyond Content Relevance: Evaluating Instruction Following in Retrieval Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=OlRjxSuSwl&name=pdf)
- Code: [https://github.com/EIT-NLP/InfoSearch](https://github.com/EIT-NLP/InfoSearch)

199、NNsight and NDIF: Democratizing Access to Open-Weight Foundation Model Internals

- Paper: [**点击下载**](https://openreview.net/attachment?id=MxbEiFRf39&name=pdf)
- Code: [https://nnsight.net/](https://nnsight.net/)

200、Robust Barycenter Estimation using Semi-Unbalanced Neural Optimal Transport

- Paper: [**点击下载**](https://openreview.net/attachment?id=CI5Cj0vktS&name=pdf)
- Code: [https://github.com/milenagazdieva/U-NOTBarycenters](https://github.com/milenagazdieva/U-NOTBarycenters)

201、Scaling up Masked Diffusion Models on Text

- Paper: [**点击下载**](https://openreview.net/attachment?id=WNvvwK0tut&name=pdf)
- Code: [https://github.com/ML-GSAI/SMDM](https://github.com/ML-GSAI/SMDM)

202、Think while You Generate: Discrete Diffusion with Planned Denoising

- Paper: [**点击下载**](https://openreview.net/attachment?id=MJNywBdSDy&name=pdf)
- Code: [https://github.com/liusulin/DDPD](https://github.com/liusulin/DDPD)

203、Evaluating Semantic Variation in Text-to-Image Synthesis: A Causal Perspective

- Paper: [**点击下载**](https://openreview.net/attachment?id=NWb128pSCb&name=pdf)
- Code: [https://github.com/zhuxiangru/SemVarBench](https://github.com/zhuxiangru/SemVarBench)

204、Can Transformers Do Enumerative Geometry?

- Paper: [**点击下载**](https://openreview.net/attachment?id=4X9RpKH4Ls&name=pdf)
- Code: [https://github.com/Baran-phys/DynamicFormer](https://github.com/Baran-phys/DynamicFormer)

205、EqNIO: Subequivariant Neural Inertial Odometry

- Paper: [**点击下载**](https://openreview.net/attachment?id=C8jXEugWkq&name=pdf)
- Code: [https://github.com/RoyinaJayanth/EqNIO](https://github.com/RoyinaJayanth/EqNIO)

206、Interaction Asymmetry: A General Principle for Learning Composable Abstractions

- Paper: [**点击下载**](https://openreview.net/attachment?id=cCl10IU836&name=pdf)
- Code: [https://www.github.com/JackBrady/interaction-asymmetry](https://www.github.com/JackBrady/interaction-asymmetry)

207、Pangea: A Fully Open Multilingual Multimodal LLM for 39 Languages

- Paper: [**点击下载**](https://openreview.net/attachment?id=a3g2l4yEys&name=pdf)
- Code: [https://neulab.github.io/Pangea/](https://neulab.github.io/Pangea/)

208、CirT: Global Subseasonal-to-Seasonal Forecasting with Geometry-inspired Transformer

- Paper: [**点击下载**](https://openreview.net/attachment?id=YslOW2SO6S&name=pdf)
- Code: [https://github.com/compasszzn/CirT](https://github.com/compasszzn/CirT)

209、Provence: efficient and robust context pruning for retrieval-augmented generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=TDy5Ih78b4&name=pdf)
- Code: [https://huggingface.co/naver/provence-reranker-debertav3-v1](https://huggingface.co/naver/provence-reranker-debertav3-v1)

210、CViT: Continuous Vision Transformer for Operator Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=cRnCcuLvyr&name=pdf)
- Code: [https://github.com/PredictiveIntelligenceLab/cvit](https://github.com/PredictiveIntelligenceLab/cvit)

211、Unveiling the Magic of Code Reasoning through Hypothesis Decomposition and Amendment

- Paper: [**点击下载**](https://openreview.net/attachment?id=kN25ggeq1J&name=pdf)
- Code: [https://github.com/TnTWoW/code_reasoning](https://github.com/TnTWoW/code_reasoning)

212、Flow: Modularized Agentic Workflow Automation

- Paper: [**点击下载**](https://openreview.net/attachment?id=sLKDbuyq99&name=pdf)
- Code: [https://github.com/tmllab/2025_ICLR_FLOW](https://github.com/tmllab/2025_ICLR_FLOW)

213、An Exploration with Entropy Constrained 3D Gaussians for 2D Video Compression

- Paper: [**点击下载**](https://openreview.net/attachment?id=JbRM5QKRDd&name=pdf)
- Code: [https://github.com/actcwlf/GSVC](https://github.com/actcwlf/GSVC)

214、Can Video LLMs Refuse to Answer? Alignment for Answerability in Video Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=P9VdRQOyqu&name=pdf)
- Code: [https://github.com/EsYoon7/UVQA](https://github.com/EsYoon7/UVQA)

215、Physics-informed Temporal Difference Metric Learning for Robot Motion Planning

- Paper: [**点击下载**](https://openreview.net/attachment?id=TOiageVNru&name=pdf)
- Code: [https://github.com/ruiqini/ntrl-demo](https://github.com/ruiqini/ntrl-demo)

216、Scalable Mechanistic Neural Networks

- Paper: [**点击下载**](https://openreview.net/attachment?id=Oazgf8A24z&name=pdf)
- Code: [https://github.com/IST-DASLab/ScalableMNN](https://github.com/IST-DASLab/ScalableMNN)

217、One Model Transfer to All: On Robust Jailbreak Prompts Generation against LLMs

- Paper: [**点击下载**](https://openreview.net/attachment?id=sULAwlAWc1&name=pdf)
- Code: [https://github.com/LLBao/ArrAttack](https://github.com/LLBao/ArrAttack)

218、E(3)-equivariant models cannot learn chirality: Field-based molecular generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=mXHTifc1Fn&name=pdf)
- Code: [https://github.com/Dumitrescu-Alexandru/FMG](https://github.com/Dumitrescu-Alexandru/FMG)

219、JPEG Inspired Deep Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=te2IdORabL&name=pdf)
- Code: [https://github.com/AhmedHussKhalifa/JPEG-Inspired-DL.git](https://github.com/AhmedHussKhalifa/JPEG-Inspired-DL.git)

220、The Case for Cleaner Biosignals: High-fidelity Neural Compressor Enables Transfer from Cleaner iEEG to Noisier EEG

- Paper: [**点击下载**](https://openreview.net/attachment?id=b57IG6N20B&name=pdf)
- Code: [https://github.com/IBM/eeg-ieeg-brain-compressor](https://github.com/IBM/eeg-ieeg-brain-compressor)

221、Tool-Planner: Task Planning with Clusters across Multiple Tools

- Paper: [**点击下载**](https://openreview.net/attachment?id=dRz3cizftU&name=pdf)
- Code: [https://github.com/OceannTwT/Tool-Planner](https://github.com/OceannTwT/Tool-Planner)

222、Systematic Relational Reasoning With Epistemic Graph Neural Networks

- Paper: [**点击下载**](https://openreview.net/attachment?id=qNp86ByQlN&name=pdf)
- Code: [https://github.com/erg0dic/gnn-sg](https://github.com/erg0dic/gnn-sg)

223、Simple, Good, Fast: Self-Supervised World Models Free of Baggage

- Paper: [**点击下载**](https://openreview.net/attachment?id=yFGR36PLDJ&name=pdf)
- Code: [https://github.com/jrobine/sgf](https://github.com/jrobine/sgf)

224、Noise-conditioned Energy-based Annealed Rewards (NEAR): A Generative Framework for Imitation Learning from Observation

- Paper: [**点击下载**](https://openreview.net/attachment?id=DL9txImSzm&name=pdf)
- Code: [https://anishhdiwan.github.io/noise-conditioned-energy-based-annealed-rewards/](https://anishhdiwan.github.io/noise-conditioned-energy-based-annealed-rewards/)

225、Towards Robust Alignment of Language Models: Distributionally Robustifying Direct Preference Optimization

- Paper: [**点击下载**](https://openreview.net/attachment?id=CbfsKHiWEn&name=pdf)
- Code: [https://github.com/junkangwu/Dr_DPO](https://github.com/junkangwu/Dr_DPO)

226、RAPID: Retrieval Augmented Training of Differentially Private Diffusion Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=txZVQRc2ab&name=pdf)
- Code: [https://github.com/TanqiuJiang/RAPID](https://github.com/TanqiuJiang/RAPID)

227、Triples as the Key: Structuring Makes Decomposition and Verification Easier in LLM-based TableQA

- Paper: [**点击下载**](https://openreview.net/attachment?id=UwcZEoNP19&name=pdf)
- Code: [https://github.com/Zhen-Yang094/TIDE-triples](https://github.com/Zhen-Yang094/TIDE-triples)

228、Agents' Room:  Narrative Generation through Multi-step Collaboration

- Paper: [**点击下载**](https://openreview.net/attachment?id=HfWcFs7XLR&name=pdf)
- Code: [https://github.com/google-deepmind/tell_me_a_story](https://github.com/google-deepmind/tell_me_a_story)

229、Equivariant Denoisers Cannot Copy Graphs: Align Your Graph Diffusion Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=onIro14tHv&name=pdf)
- Code: [https://github.com/Aalto-QuML/DiffAlign](https://github.com/Aalto-QuML/DiffAlign)

230、MrSteve: Instruction-Following Agents in Minecraft with What-Where-When Memory

- Paper: [**点击下载**](https://openreview.net/attachment?id=CjXaMI2kUH&name=pdf)
- Code: [https://sites.google.com/view/mr-steve](https://sites.google.com/view/mr-steve)

231、Semi-Supervised Vision-Centric 3D Occupancy World Model for Autonomous Driving

- Paper: [**点击下载**](https://openreview.net/attachment?id=rCX9l4OTCT&name=pdf)
- Code: [https://github.com/getterupper/PreWorld](https://github.com/getterupper/PreWorld)

232、InstantSplamp: Fast and Generalizable Stenography Framework for Generative Gaussian Splatting

- Paper: [**点击下载**](https://openreview.net/attachment?id=xvhV3LvYTc&name=pdf)
- Code: [https://gaussian-stego.github.io/](https://gaussian-stego.github.io/)

233、Facilitating Multi-turn Function Calling for LLMs via Compositional Instruction Tuning

- Paper: [**点击下载**](https://openreview.net/attachment?id=owP2mymrTD&name=pdf)
- Code: [https://github.com/PKU-Baichuan-MLSystemLab/BUTTON](https://github.com/PKU-Baichuan-MLSystemLab/BUTTON)

234、COFlowNet: Conservative Constraints on Flows Enable High-Quality Candidate Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=tXUkT709OJ&name=pdf)
- Code: [https://github.com/yuxuan9982/COflownet](https://github.com/yuxuan9982/COflownet)

235、Multimodal Quantitative Language for Generative Recommendation

- Paper: [**点击下载**](https://openreview.net/attachment?id=v7YrIjpkTF&name=pdf)
- Code: [https://github.com/zhaijianyang/MQL4GRec](https://github.com/zhaijianyang/MQL4GRec)

236、Monet: Mixture of Monosemantic Experts for Transformers

- Paper: [**点击下载**](https://openreview.net/attachment?id=1Ogw1SHY3p&name=pdf)
- Code: [https://github.com/dmis-lab/Monet](https://github.com/dmis-lab/Monet)

237、Graph-based Document Structure Analysis

- Paper: [**点击下载**](https://openreview.net/attachment?id=Fu0aggezN9&name=pdf)
- Code: [https://yufanchen96.github.io/projects/GraphDoc](https://yufanchen96.github.io/projects/GraphDoc)

238、Dream to Manipulate: Compositional World Models Empowering Robot Imitation Learning with Imagination

- Paper: [**点击下载**](https://openreview.net/attachment?id=3RSLW9YSgk&name=pdf)
- Code: [https://dreamtomanipulate.github.io/](https://dreamtomanipulate.github.io/)

239、3D-MolT5: Leveraging Discrete Structural Information for Molecule-Text Modeling

- Paper: [**点击下载**](https://openreview.net/attachment?id=eGqQyTAbXC&name=pdf)
- Code: [https://github.com/QizhiPei/3D-MolT5](https://github.com/QizhiPei/3D-MolT5)

240、Redefining the task of Bioactivity Prediction

- Paper: [**点击下载**](https://openreview.net/attachment?id=S8gbnkCgxZ&name=pdf)
- Code: [https://github.com/bowen-gao/SIU](https://github.com/bowen-gao/SIU)

241、Robust Simulation-Based Inference under Missing Data via Neural Processes

- Paper: [**点击下载**](https://openreview.net/attachment?id=GsR3zRCRX5&name=pdf)
- Code: [https://github.com/Aalto-QuML/RISE](https://github.com/Aalto-QuML/RISE)

242、Physics-Informed Diffusion Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=tpYeermigp&name=pdf)
- Code: [https://github.com/jhbastek/PhysicsInformedDiffusionModels](https://github.com/jhbastek/PhysicsInformedDiffusionModels)

243、DynFrs: An Efficient Framework for Machine Unlearning in Random Forest

- Paper: [**点击下载**](https://openreview.net/attachment?id=nsCOeCLR8e&name=pdf)
- Code: [shurong.22](shurong.22)

244、Mix-LN: Unleashing the Power of Deeper Layers by Combining Pre-LN and Post-LN

- Paper: [**点击下载**](https://openreview.net/attachment?id=BChpQU64RG&name=pdf)
- Code: [https://github.com/pixeli99/MixLN](https://github.com/pixeli99/MixLN)

245、Decoupling Layout from Glyph in Online Chinese Handwriting Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=DhHIw9Nbl1&name=pdf)
- Code: [https://github.com/singularityrms/OLHWG](https://github.com/singularityrms/OLHWG)

246、Diversity Empowers Intelligence: Integrating Expertise of Software Engineering Agents

- Paper: [**点击下载**](https://openreview.net/attachment?id=cKlzKs3Nnb&name=pdf)
- Code: [https://github.com/SalesforceAIResearch/swecomm](https://github.com/SalesforceAIResearch/swecomm)

247、ADBM: Adversarial Diffusion Bridge Model for Reliable Adversarial Purification

- Paper: [**点击下载**](https://openreview.net/attachment?id=g0rnZeBguq&name=pdf)
- Code: [https://github.com/LixiaoTHU/ADBM](https://github.com/LixiaoTHU/ADBM)

248、Divide and Translate: Compositional First-Order Logic Translation and Verification for Complex Logical Reasoning

- Paper: [**点击下载**](https://openreview.net/attachment?id=09FiNmvNMw&name=pdf)
- Code: [https://github.com/Hyun-Ryu/clover](https://github.com/Hyun-Ryu/clover)

249、Pursuing Feature Separation based on Neural Collapse for Out-of-Distribution Detection

- Paper: [**点击下载**](https://openreview.net/attachment?id=mUXdysoxEP&name=pdf)
- Code: [https://github.com/Wuyingwen/Pursuing-Feature-Separation-for-OOD-Detection](https://github.com/Wuyingwen/Pursuing-Feature-Separation-for-OOD-Detection)

250、MA$^2$E: Addressing Partial Observability in Multi-Agent Reinforcement Learning with Masked Auto-Encoder

- Paper: [**点击下载**](https://openreview.net/attachment?id=klpdEThT8q&name=pdf)
- Code: [https://github.com/cheesebro329/MA2E](https://github.com/cheesebro329/MA2E)

251、Wayward Concepts In Large Multimodal Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=74vnDs1R97&name=pdf)
- Code: [https://wayward-concepts.github.io](https://wayward-concepts.github.io)

252、Towards Hierarchical Rectified Flow

- Paper: [**点击下载**](https://openreview.net/attachment?id=6F6qwdycgJ&name=pdf)
- Code: [https://riccizz.github.io/HRF/](https://riccizz.github.io/HRF/)

253、Zeroth-Order Fine-Tuning of LLMs with Transferable Static Sparsity

- Paper: [**点击下载**](https://openreview.net/attachment?id=myYzr50xBh&name=pdf)
- Code: [https://github.com/GarlGuo/SensZOQ](https://github.com/GarlGuo/SensZOQ)

254、MMR: A Large-scale Benchmark Dataset for Multi-target and Multi-granularity Reasoning Segmentation

- Paper: [**点击下载**](https://openreview.net/attachment?id=mzL19kKE3r&name=pdf)
- Code: [https://github.com/jdg900/MMR](https://github.com/jdg900/MMR)

255、Do Egocentric Video-Language Models Truly Understand Hand-Object Interactions?

- Paper: [**点击下载**](https://openreview.net/attachment?id=M8gXSFGkn2&name=pdf)
- Code: [https://github.com/xuboshen/EgoNCEpp](https://github.com/xuboshen/EgoNCEpp)

256、The "Law'' of the Unconscious Contrastive Learner: Probabilistic Alignment of Unpaired Modalities

- Paper: [**点击下载**](https://openreview.net/attachment?id=DsIOUoZkVk&name=pdf)
- Code: [https://github.com/YongweiChe/UnconsciousContrastiveLearner](https://github.com/YongweiChe/UnconsciousContrastiveLearner)

257、System 1.x: Learning to Balance Fast and Slow Planning with Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=zd0iX5xBhA&name=pdf)
- Code: [https://github.com/swarnaHub/System-1.x](https://github.com/swarnaHub/System-1.x)

258、Investigating the Pre-Training Dynamics of In-Context Learning: Task Recognition vs. Task Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=htDczodFN5&name=pdf)
- Code: [https://github.com/RUCAIBox/Competitive-ICL](https://github.com/RUCAIBox/Competitive-ICL)

259、Advancing Graph Generation through Beta Diffusion

- Paper: [**点击下载**](https://openreview.net/attachment?id=x1An5a3U9I&name=pdf)
- Code: [https://github.com/xinyangATK/GraphBetaDiffusion](https://github.com/xinyangATK/GraphBetaDiffusion)

260、UniWav: Towards Unified Pre-training for Speech Representation Learning and Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=yj9lLwMjnE&name=pdf)
- Code: [https://alexander-h-liu.github.io/uniwav-demo.github.io/](https://alexander-h-liu.github.io/uniwav-demo.github.io/)

261、Multiple Heads are Better than One: Mixture of Modality Knowledge Experts for Entity Representation Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=ue1Tt3h1VC&name=pdf)
- Code: [https://github.com/zjukg/MoMoK](https://github.com/zjukg/MoMoK)

262、Confidence Elicitation: A New Attack Vector for Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=aTYexOYlLb&name=pdf)
- Code: [https://github.com/Aniloid2/Confidence_Elicitation_Attacks](https://github.com/Aniloid2/Confidence_Elicitation_Attacks)

263、CoMRes: Semi-Supervised Time Series Forecasting Utilizing Consensus Promotion of Multi-Resolution

- Paper: [**点击下载**](https://openreview.net/attachment?id=bRa4JLPzii&name=pdf)
- Code: [https://github.com/yjucho1/CoMRes](https://github.com/yjucho1/CoMRes)

264、MCNC: Manifold-Constrained Reparameterization for Neural Compression

- Paper: [**点击下载**](https://openreview.net/attachment?id=VMV8gefvq8&name=pdf)
- Code: [https://github.com/mint-vu/MCNC](https://github.com/mint-vu/MCNC)

265、HARDMath: A Benchmark Dataset for Challenging Problems in Applied Mathematics

- Paper: [**点击下载**](https://openreview.net/attachment?id=nDTvP6tBMd&name=pdf)
- Code: [https://github.com/sarahmart/HARDMath](https://github.com/sarahmart/HARDMath)

266、Not All Heads Matter: A Head-Level KV Cache Compression Method with Integrated Retrieval and Reasoning

- Paper: [**点击下载**](https://openreview.net/attachment?id=FJFVmeXusW&name=pdf)
- Code: [https://github.com/FYYFU/HeadKV](https://github.com/FYYFU/HeadKV)

267、AdaIR: Adaptive All-in-One Image Restoration via Frequency Mining and Modulation

- Paper: [**点击下载**](https://openreview.net/attachment?id=M5t0WvjfCg&name=pdf)
- Code: [https://github.com/c-yn/AdaIR](https://github.com/c-yn/AdaIR)

268、Forte : Finding Outliers with Representation Typicality Estimation

- Paper: [**点击下载**](https://openreview.net/attachment?id=7XNgVPxCiA&name=pdf)
- Code: [https://github.com/DebarghaG/forte](https://github.com/DebarghaG/forte)

269、LLM-wrapper: Black-Box Semantic-Aware Adaptation of Vision-Language Models for Referring Expression Comprehension

- Paper: [**点击下载**](https://openreview.net/attachment?id=PgXpOOqtyd&name=pdf)
- Code: [https://github.com/valeoai/LLM_wrapper](https://github.com/valeoai/LLM_wrapper)

270、NextBestPath: Efficient 3D Mapping of Unseen Environments

- Paper: [**点击下载**](https://openreview.net/attachment?id=7WaRh4gCXp&name=pdf)
- Code: [https://shiyao-li.github.io/nbp/](https://shiyao-li.github.io/nbp/)

271、LLaRA: Supercharging Robot Learning Data for Vision-Language Policy

- Paper: [**点击下载**](https://openreview.net/attachment?id=iVxxgZlXh6&name=pdf)
- Code: [https://github.com/LostXine/LLaRA](https://github.com/LostXine/LLaRA)

272、Breaking the Reclustering Barrier in Centroid-based Deep Clustering

- Paper: [**点击下载**](https://openreview.net/attachment?id=r01fcKhzT5&name=pdf)
- Code: [https://github.com/Probabilistic-and-Interactive-ML/breaking-the-reclustering-barrier](https://github.com/Probabilistic-and-Interactive-ML/breaking-the-reclustering-barrier)

273、Closed-Form Merging of Parameter-Efficient Modules for Federated Continual Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=ROpY0qRUXL&name=pdf)
- Code: [https://github.com/aimagelab/fed-mammoth](https://github.com/aimagelab/fed-mammoth)

274、Frequency-Guided Masking for Enhanced Vision Self-Supervised Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=VmJdqhuTCh&name=pdf)
- Code: [https://github.com/aminK8/FOLK](https://github.com/aminK8/FOLK)

275、Generalized Consistency Trajectory Models for Image Manipulation

- Paper: [**点击下载**](https://openreview.net/attachment?id=Zjv38dg1Hb&name=pdf)
- Code: [https://github.com/1202kbs/GCTM](https://github.com/1202kbs/GCTM)

276、API Pack: A Massive Multi-Programming Language Dataset for API Call Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=f7O3hITh5s&name=pdf)
- Code: [https://github.com/zguo0525/API-Pack](https://github.com/zguo0525/API-Pack)

277、Spherical Tree-Sliced Wasserstein Distance

- Paper: [**点击下载**](https://openreview.net/attachment?id=FPQzXME9NK&name=pdf)
- Code: [https://github.com/lilythchu/STSW.git](https://github.com/lilythchu/STSW.git)

278、Dualformer: Controllable Fast and Slow Thinking by Learning with Randomized Reasoning Traces

- Paper: [**点击下载**](https://openreview.net/attachment?id=bmbRCRiNDu&name=pdf)
- Code: [https://github.com/facebookresearch/dualformer](https://github.com/facebookresearch/dualformer)

279、Training Robust Ensembles Requires Rethinking Lipschitz Continuity

- Paper: [**点击下载**](https://openreview.net/attachment?id=WKW5TG8ItY&name=pdf)
- Code: [https://github.com/Ali-E/LOTOS](https://github.com/Ali-E/LOTOS)

280、Sparse Autoencoders Do Not Find Canonical Units of Analysis

- Paper: [**点击下载**](https://openreview.net/attachment?id=9ca9eHNrdH&name=pdf)
- Code: [https://metasaes.streamlit.app/](https://metasaes.streamlit.app/)

281、MUSE: Machine Unlearning Six-Way Evaluation for Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=TArmA033BU&name=pdf)
- Code: [https://muse-bench.github.io](https://muse-bench.github.io)

282、Learning Structured Representations by Embedding Class Hierarchy with Fast Optimal Transport

- Paper: [**点击下载**](https://openreview.net/attachment?id=AnL6BuWzxa&name=pdf)
- Code: [https://github.com/uiuctml/OTCPCC](https://github.com/uiuctml/OTCPCC)

283、Small Models are LLM Knowledge Triggers for Medical Tabular Prediction

- Paper: [**点击下载**](https://openreview.net/attachment?id=WoPovNkM5h&name=pdf)
- Code: [https://github.com/jyansir/sersal](https://github.com/jyansir/sersal)

284、PQMass: Probabilistic Assessment of the Quality of Generative Models using Probability Mass Estimation

- Paper: [**点击下载**](https://openreview.net/attachment?id=n7qGCmluZr&name=pdf)
- Code: [https://github.com/Ciela-Institute/PQM](https://github.com/Ciela-Institute/PQM)

285、Simulating Training Dynamics to Reconstruct Training Data from Deep Neural Networks

- Paper: [**点击下载**](https://openreview.net/attachment?id=ZJftXKy12x&name=pdf)
- Code: [https://github.com/BlueBlood6/SimuDy](https://github.com/BlueBlood6/SimuDy)

286、BANGS: Game-theoretic Node Selection for Graph Self-Training

- Paper: [**点击下载**](https://openreview.net/attachment?id=h51mpl8Tyx&name=pdf)
- Code: [https://github.com/fangxin-wang/BANGS](https://github.com/fangxin-wang/BANGS)

287、UNSURE: self-supervised learning with Unknown Noise level  and Stein's Unbiased Risk Estimate

- Paper: [**点击下载**](https://openreview.net/attachment?id=ScVnYBaSEw&name=pdf)
- Code: [https://github.com/tachella/unsure](https://github.com/tachella/unsure)

288、Generating Likely Counterfactuals Using Sum-Product Networks

- Paper: [**点击下载**](https://openreview.net/attachment?id=rGyi8NNqB0&name=pdf)
- Code: [https://github.com/Epanemu/LiCE](https://github.com/Epanemu/LiCE)

289、Unintentional Unalignment: Likelihood Displacement in Direct Preference Optimization

- Paper: [**点击下载**](https://openreview.net/attachment?id=uaMSBJDnRv&name=pdf)
- Code: [https://github.com/princeton-nlp/unintentional-unalignment](https://github.com/princeton-nlp/unintentional-unalignment)

290、PooDLe🐩: Pooled and dense self-supervised learning from naturalistic videos

- Paper: [**点击下载**](https://openreview.net/attachment?id=dEg5SdGaiq&name=pdf)
- Code: [https://agenticlearning.ai/poodle/](https://agenticlearning.ai/poodle/)

291、TweedieMix: Improving Multi-Concept Fusion for Diffusion-based Image/Video Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=ee2c4MEx9l&name=pdf)
- Code: [https://github.com/KwonGihyun/TweedieMix](https://github.com/KwonGihyun/TweedieMix)

292、CHiP: Cross-modal Hierarchical Direct Preference Optimization for Multimodal LLMs

- Paper: [**点击下载**](https://openreview.net/attachment?id=7lpDn2MhM2&name=pdf)
- Code: [https://github.com/LVUGAI/CHiP](https://github.com/LVUGAI/CHiP)

293、TD-Paint: Faster Diffusion Inpainting Through Time Aware Pixel Conditioning

- Paper: [**点击下载**](https://openreview.net/attachment?id=erWwBoR59l&name=pdf)
- Code: [https://github.com/MaugrimEP/td-paint](https://github.com/MaugrimEP/td-paint)

294、Number Cookbook: Number Understanding of Language Models and How to Improve It

- Paper: [**点击下载**](https://openreview.net/attachment?id=BWS5gVjgeY&name=pdf)
- Code: [https://github.com/GraphPKU/number_cookbook](https://github.com/GraphPKU/number_cookbook)

295、Bringing NeRFs to the Latent Space: Inverse Graphics Autoencoder

- Paper: [**点击下载**](https://openreview.net/attachment?id=LTDtjrv02Y&name=pdf)
- Code: [https://ig-ae.github.io](https://ig-ae.github.io)

296、Agent Security Bench (ASB): Formalizing and Benchmarking Attacks and Defenses in LLM-based Agents

- Paper: [**点击下载**](https://openreview.net/attachment?id=V4y0CpX4hK&name=pdf)
- Code: [https://github.com/agiresearch/ASB](https://github.com/agiresearch/ASB)

297、Factor Graph-based Interpretable Neural Networks

- Paper: [**点击下载**](https://openreview.net/attachment?id=10DtLPsdro&name=pdf)
- Code: [https://github.com/yushuowiki/AGAIN](https://github.com/yushuowiki/AGAIN)

298、Alchemy: Amplifying Theorem-Proving Capability Through Symbolic Mutation

- Paper: [**点击下载**](https://openreview.net/attachment?id=7NL74jUiMg&name=pdf)
- Code: [https://github.com/wclsn/Alchemy](https://github.com/wclsn/Alchemy)

299、Agent Skill Acquisition for Large Language Models via CycleQD

- Paper: [**点击下载**](https://openreview.net/attachment?id=Kvdh12wGC0&name=pdf)
- Code: [https://github.com/SakanaAI/CycleQD](https://github.com/SakanaAI/CycleQD)

300、KBLaM: Knowledge Base augmented Language Model

- Paper: [**点击下载**](https://openreview.net/attachment?id=aLsMzkTej9&name=pdf)
- Code: [https://github.com/microsoft/KBLaM/](https://github.com/microsoft/KBLaM/)

301、MOFFlow: Flow Matching for Structure Prediction of Metal-Organic Frameworks

- Paper: [**点击下载**](https://openreview.net/attachment?id=dNT3abOsLo&name=pdf)
- Code: [https://github.com/nayoung10/MOFFlow](https://github.com/nayoung10/MOFFlow)

302、An Information Criterion for Controlled Disentanglement of Multimodal Data

- Paper: [**点击下载**](https://openreview.net/attachment?id=3n4RY25UWP&name=pdf)
- Code: [https://github.com/uhlerlab/DisentangledSSL](https://github.com/uhlerlab/DisentangledSSL)

303、ReAttention: Training-Free Infinite Context with Finite Attention Scope

- Paper: [**点击下载**](https://openreview.net/attachment?id=KDGP8yAz5b&name=pdf)
- Code: [https://github.com/OpenMOSS/ReAttention](https://github.com/OpenMOSS/ReAttention)

304、SORRY-Bench: Systematically Evaluating Large Language Model Safety Refusal

- Paper: [**点击下载**](https://openreview.net/attachment?id=YfKNaRktan&name=pdf)
- Code: [https://sorry-bench.github.io](https://sorry-bench.github.io)

305、Periodic Materials Generation using Text-Guided Joint Diffusion Model

- Paper: [**点击下载**](https://openreview.net/attachment?id=AkBrb7yQ0G&name=pdf)
- Code: [https://github.com/kdmsit/TGDMat](https://github.com/kdmsit/TGDMat)

306、QERA: an Analytical Framework for Quantization Error Reconstruction

- Paper: [**点击下载**](https://openreview.net/attachment?id=LB5cKhgOTu&name=pdf)
- Code: [https://github.com/ChengZhang-98/QERA](https://github.com/ChengZhang-98/QERA)

307、Efficient Discovery of Pareto Front for Multi-Objective Reinforcement Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=fDGPIuCdGi&name=pdf)
- Code: [https://github.com/RuohLiuq/C-MORL](https://github.com/RuohLiuq/C-MORL)

308、Semi-Supervised CLIP Adaptation by Enforcing Semantic and Trapezoidal Consistency

- Paper: [**点击下载**](https://openreview.net/attachment?id=97D725GJtQ&name=pdf)
- Code: [https://github.com/Gank0078/SemiCLIP](https://github.com/Gank0078/SemiCLIP)

309、Unsupervised Zero-Shot Reinforcement Learning via Dual-Value Forward-Backward Representation

- Paper: [**点击下载**](https://openreview.net/attachment?id=0QnKnt411O&name=pdf)
- Code: [https://github.com/bofusun/DVFB](https://github.com/bofusun/DVFB)

310、Predicting the Energy Landscape of Stochastic Dynamical System via  Physics-informed Self-supervised Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=PxRATSTDlS&name=pdf)
- Code: [https://github.com/tsinghua-fib-lab/PESLA](https://github.com/tsinghua-fib-lab/PESLA)

311、Biologically Plausible Brain Graph Transformer

- Paper: [**点击下载**](https://openreview.net/attachment?id=rQyg6MnsDb&name=pdf)
- Code: [https://github.com/pcyyyy/BioBGT](https://github.com/pcyyyy/BioBGT)

312、Inference Scaling Laws: An Empirical Analysis of Compute-Optimal Inference for LLM Problem-Solving

- Paper: [**点击下载**](https://openreview.net/attachment?id=VNckp7JEHn&name=pdf)
- Code: [https://thu-wyz.github.io/inference-scaling/](https://thu-wyz.github.io/inference-scaling/)

313、UV-Attack: Physical-World Adversarial Attacks on Person Detection via Dynamic-NeRF-based UV Mapping

- Paper: [**点击下载**](https://openreview.net/attachment?id=pqeWzZTrZY&name=pdf)
- Code: [https://github.com/PolyLiYJ/UV-Attack](https://github.com/PolyLiYJ/UV-Attack)

314、Breaking Free from MMI: A New Frontier in Rationalization by Probing Input Utilization

- Paper: [**点击下载**](https://openreview.net/attachment?id=WZ0s2smcKP&name=pdf)
- Code: [https://github.com/jugechengzi/Rationalization-N2R](https://github.com/jugechengzi/Rationalization-N2R)

315、Optimistic Games for Combinatorial Bayesian Optimization with Application to Protein Design

- Paper: [**点击下载**](https://openreview.net/attachment?id=xiyzCfXTS6&name=pdf)
- Code: [https://github.com/melisilaydabal/gameopt](https://github.com/melisilaydabal/gameopt)

316、DAWN: Dynamic Frame Avatar with Non-autoregressive Diffusion Framework for Talking head Video Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=vjHySpxDsv&name=pdf)
- Code: [https://hanbo-cheng.github.io/DAWN](https://hanbo-cheng.github.io/DAWN)

317、Vision and Language Synergy for Rehearsal Free Continual Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=9aZ2ixiYGd&name=pdf)
- Code: [https://github.com/anwarmaxsum/LEAPGEN](https://github.com/anwarmaxsum/LEAPGEN)

318、Rethinking Multiple-Instance Learning From Feature Space to Probability Space

- Paper: [**点击下载**](https://openreview.net/attachment?id=torbeUlslS&name=pdf)
- Code: [https://github.com/LMBDA-design/PSAMIL](https://github.com/LMBDA-design/PSAMIL)

319、Why Does the Effective Context Length of LLMs Fall Short?

- Paper: [**点击下载**](https://openreview.net/attachment?id=eoln5WgrPx&name=pdf)
- Code: [https://github.com/HKUNLP/STRING](https://github.com/HKUNLP/STRING)

320、Complementary Label Learning with Positive Label Guessing and Negative Label Enhancement

- Paper: [**点击下载**](https://openreview.net/attachment?id=LPRxGZ7Oax&name=pdf)
- Code: [https://github.com/yhli-ml/PLNL](https://github.com/yhli-ml/PLNL)

321、Training-free LLM-generated Text Detection by Mining Token Probability Sequences

- Paper: [**点击下载**](https://openreview.net/attachment?id=vo4AHjowKi&name=pdf)
- Code: [https://github.com/TrustMedia-zju/Lastde_Detector](https://github.com/TrustMedia-zju/Lastde_Detector)

322、Dataset Ownership Verification in Contrastive Pre-trained Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=zeAOzn80VQ&name=pdf)
- Code: [https://github.com/xieyc99/DOV4CL](https://github.com/xieyc99/DOV4CL)

323、MMFakeBench: A Mixed-Source Multimodal Misinformation Detection Benchmark for LVLMs

- Paper: [**点击下载**](https://openreview.net/attachment?id=D6zn6ozJs7&name=pdf)
- Code: [https://liuxuannan.github.io/MMFakeBench.github.io/](https://liuxuannan.github.io/MMFakeBench.github.io/)

324、Standardizing Structural Causal Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=aXuWowhIYt&name=pdf)
- Code: [https://github.com/werkaaa/iscm](https://github.com/werkaaa/iscm)

325、OpenVid-1M: A Large-Scale High-Quality Dataset for Text-to-video Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=j7kdXSrISM&name=pdf)
- Code: [https://nju-pcalab.github.io/projects/openvid](https://nju-pcalab.github.io/projects/openvid)

326、VisRAG: Vision-based Retrieval-augmented Generation on Multi-modality Documents

- Paper: [**点击下载**](https://openreview.net/attachment?id=zG459X3Xge&name=pdf)
- Code: [https://github.com/openbmb/visrag](https://github.com/openbmb/visrag)

327、Bootstrapping Language Models with DPO Implicit Rewards

- Paper: [**点击下载**](https://openreview.net/attachment?id=dliIIodM6b&name=pdf)
- Code: [https://github.com/sail-sg/dice](https://github.com/sail-sg/dice)

328、Learning system dynamics without forgetting

- Paper: [**点击下载**](https://openreview.net/attachment?id=rjuZyMfLSd&name=pdf)
- Code: [https://github.com/QueuQ/MS-GODE](https://github.com/QueuQ/MS-GODE)

329、Reconsidering Faithfulness in Regular, Self-Explainable and Domain Invariant GNNs

- Paper: [**点击下载**](https://openreview.net/attachment?id=kiOxNsrpQy&name=pdf)
- Code: [https://github.com/steveazzolin/reconsidering-faithfulness-in-gnns](https://github.com/steveazzolin/reconsidering-faithfulness-in-gnns)

330、SWIFT: On-the-Fly Self-Speculative Decoding for LLM Inference Acceleration

- Paper: [**点击下载**](https://openreview.net/attachment?id=EKJhH5D5wA&name=pdf)
- Code: [https://github.com/hemingkx/SWIFT](https://github.com/hemingkx/SWIFT)

331、SysBench: Can LLMs Follow System Message?

- Paper: [**点击下载**](https://openreview.net/attachment?id=KZWaxtzIRx&name=pdf)
- Code: [https://github.com/PKU-Baichuan-MLSystemLab/SysBench](https://github.com/PKU-Baichuan-MLSystemLab/SysBench)

332、Fast Direct: Query-Efficient  Online Black-box Guidance  for Diffusion-model Target Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=OmpTdjl7RV&name=pdf)
- Code: [https://github.com/kimyong95/guide-stable-diffusion/tree/fast-direct](https://github.com/kimyong95/guide-stable-diffusion/tree/fast-direct)

333、AdaRankGrad: Adaptive Gradient Rank and Moments for Memory-Efficient LLMs Training and Fine-Tuning

- Paper: [**点击下载**](https://openreview.net/attachment?id=LvNROciCne&name=pdf)
- Code: [https://github.com/jsvir/AdaRankGrad](https://github.com/jsvir/AdaRankGrad)

334、OmniBind: Large-scale Omni Multimodal Representation via Binding Spaces

- Paper: [**点击下载**](https://openreview.net/attachment?id=l2izo0z7gu&name=pdf)
- Code: [http://omnibind.github.io](http://omnibind.github.io)

335、Arithmetic Without Algorithms: Language Models Solve Math with a Bag of Heuristics

- Paper: [**点击下载**](https://openreview.net/attachment?id=O9YTt26r2P&name=pdf)
- Code: [https://github.com/technion-cs-nlp/llm-arithmetic-heuristics](https://github.com/technion-cs-nlp/llm-arithmetic-heuristics)

336、Equivariant Neural Functional Networks for Transformers

- Paper: [**点击下载**](https://openreview.net/attachment?id=uBai0ukstY&name=pdf)
- Code: [https://github.com/MathematicalAI-NUS/Transformer-NFN](https://github.com/MathematicalAI-NUS/Transformer-NFN)

337、Weighted-Reward Preference Optimization for Implicit Model Fusion

- Paper: [**点击下载**](https://openreview.net/attachment?id=fq24pEb8SL&name=pdf)
- Code: [https://github.com/SLIT-AI/WRPO](https://github.com/SLIT-AI/WRPO)

338、Locality-aware Gaussian Compression for Fast and High-quality Rendering

- Paper: [**点击下载**](https://openreview.net/attachment?id=dHYwfV2KeP&name=pdf)
- Code: [https://seungjooshin.github.io/LocoGS](https://seungjooshin.github.io/LocoGS)

339、A Large-scale Dataset and Benchmark for Commuting Origin-Destination Flow Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=WeJEidTzff&name=pdf)
- Code: [https://github.com/tsinghua-fib-lab/CommutingODGen-Dataset](https://github.com/tsinghua-fib-lab/CommutingODGen-Dataset)

340、NeRAF: 3D Scene Infused Neural Radiance and Acoustic Fields

- Paper: [**点击下载**](https://openreview.net/attachment?id=njvSBvtiwp&name=pdf)
- Code: [https://amandinebtto.github.io/NeRAF](https://amandinebtto.github.io/NeRAF)

341、InstaRevive: One-Step Image Enhancement via Dynamic Score Matching

- Paper: [**点击下载**](https://openreview.net/attachment?id=G1CN7R5qwE&name=pdf)
- Code: [https://github.com/EternalEvan/InstaRevive](https://github.com/EternalEvan/InstaRevive)

342、A Distributional Approach to Uncertainty-Aware Preference Alignment Using Offline Demonstrations

- Paper: [**点击下载**](https://openreview.net/attachment?id=RKOAU5ti1y&name=pdf)
- Code: [https://github.com/Jasonxu1225/UA-PbRL](https://github.com/Jasonxu1225/UA-PbRL)

343、Unsupervised Multiple Kernel Learning for Graphs via Ordinality Preservation

- Paper: [**点击下载**](https://openreview.net/attachment?id=6nb2J90XJD&name=pdf)
- Code: [https://github.com/yan-sun-x/Ensemble_Kernel/](https://github.com/yan-sun-x/Ensemble_Kernel/)

344、MolSpectra: Pre-training 3D Molecular Representation with Multi-modal Energy Spectra

- Paper: [**点击下载**](https://openreview.net/attachment?id=xJDxVDG3x2&name=pdf)
- Code: [https://github.com/AzureLeon1/MolSpectra](https://github.com/AzureLeon1/MolSpectra)

345、Precise Localization of Memories: A Fine-grained Neuron-level Knowledge Editing Technique for LLMs

- Paper: [**点击下载**](https://openreview.net/attachment?id=5xP1HDvpXI&name=pdf)
- Code: [https://github.com/opanhw/FiNE](https://github.com/opanhw/FiNE)

346、Towards a Theoretical Understanding of Synthetic Data in LLM Post-Training: A Reverse-Bottleneck Perspective

- Paper: [**点击下载**](https://openreview.net/attachment?id=UxkznlcnHf&name=pdf)
- Code: [https://github.com/ZyGan1999/Towards-a-Theoretical-Understanding-of-Synthetic-Data-in-LLM-Post-Training](https://github.com/ZyGan1999/Towards-a-Theoretical-Understanding-of-Synthetic-Data-in-LLM-Post-Training)

347、On the Performance Analysis of Momentum Method: A Frequency Domain Perspective

- Paper: [**点击下载**](https://openreview.net/attachment?id=tznvtmSEiN&name=pdf)
- Code: [https://github.com/yinleung/FSGDM](https://github.com/yinleung/FSGDM)

348、Learning to Contextualize Web Pages for Enhanced Decision Making by LLM Agents

- Paper: [**点击下载**](https://openreview.net/attachment?id=3Gzz7ZQLiz&name=pdf)
- Code: [https://lcowiclr2025.github.io](https://lcowiclr2025.github.io)

349、Jamba: Hybrid Transformer-Mamba Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=JFPaD7lpBD&name=pdf)
- Code: [https://huggingface.co/ai21labs](https://huggingface.co/ai21labs)

350、Learning 3D Perception from Others' Predictions

- Paper: [**点击下载**](https://openreview.net/attachment?id=Ylk98vWQuQ&name=pdf)
- Code: [https://jinsuyoo.info/rnb-pop](https://jinsuyoo.info/rnb-pop)

351、Exploring Prosocial Irrationality for LLM Agents: A Social Cognition View

- Paper: [**点击下载**](https://openreview.net/attachment?id=u8VOQVzduP&name=pdf)
- Code: [https://github.com/XuanL17/CogMir](https://github.com/XuanL17/CogMir)

352、Incorporating Visual Correspondence into Diffusion Model for Virtual Try-On

- Paper: [**点击下载**](https://openreview.net/attachment?id=XXzOzJRyOZ&name=pdf)
- Code: [https://github.com/HiDream-ai/SPM-Diff](https://github.com/HiDream-ai/SPM-Diff)

353、Universal Image Restoration Pre-training via Degradation Classification

- Paper: [**点击下载**](https://openreview.net/attachment?id=PacBhLzeGO&name=pdf)
- Code: [https://github.com/MILab-PKU/dcpt](https://github.com/MILab-PKU/dcpt)

354、ConMix: Contrastive Mixup at Representation Level for Long-tailed Deep Clustering

- Paper: [**点击下载**](https://openreview.net/attachment?id=3lH8WT0fhu&name=pdf)
- Code: [https://github.com/LZX-001/ConMix](https://github.com/LZX-001/ConMix)

355、ADePT: Adaptive Decomposed Prompt Tuning for Parameter-Efficient Fine-tuning

- Paper: [**点击下载**](https://openreview.net/attachment?id=fswihJIYbd&name=pdf)
- Code: [https://github.com/HungerPWAY/ADePT](https://github.com/HungerPWAY/ADePT)

356、Designing Concise ConvNets with Columnar Stages

- Paper: [**点击下载**](https://openreview.net/attachment?id=zvaiz3FjA9&name=pdf)
- Code: [https://github.com/ashishkumar822/CoSNet](https://github.com/ashishkumar822/CoSNet)

357、Taming Overconfidence in LLMs: Reward Calibration in RLHF

- Paper: [**点击下载**](https://openreview.net/attachment?id=l0tg0jzsdL&name=pdf)
- Code: [https://github.com/SeanLeng1/Reward-Calibration](https://github.com/SeanLeng1/Reward-Calibration)

358、Transition Path Sampling with Improved Off-Policy Training of Diffusion Path Samplers

- Paper: [**点击下载**](https://openreview.net/attachment?id=WQV9kB1qSU&name=pdf)
- Code: [https://kiyoung98.github.io/tps-dps/](https://kiyoung98.github.io/tps-dps/)

359、StringLLM: Understanding the String Processing Capability of Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=kTXChtaaNO&name=pdf)
- Code: [https://github.com/wxl-lxw/StringLLM](https://github.com/wxl-lxw/StringLLM)

360、Consistent Flow Distillation for Text-to-3D Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=A51NEXIq1J&name=pdf)
- Code: [https://runjie-yan.github.io/cfd/](https://runjie-yan.github.io/cfd/)

361、A Benchmark for Semantic Sensitive Information in LLMs Outputs

- Paper: [**点击下载**](https://openreview.net/attachment?id=p3mxzKmuZy&name=pdf)
- Code: [https://semsi-project.github.io/](https://semsi-project.github.io/)

362、xFinder: Large Language Models as Automated Evaluators for Reliable Evaluation

- Paper: [**点击下载**](https://openreview.net/attachment?id=7UqQJUKaLM&name=pdf)
- Code: [https://github.com/IAAR-Shanghai/xFinder](https://github.com/IAAR-Shanghai/xFinder)

363、From Commands to Prompts: LLM-based Semantic File System for AIOS

- Paper: [**点击下载**](https://openreview.net/attachment?id=2G021ZqUEZ&name=pdf)
- Code: [https://github.com/agiresearch/AIOS-LSFS](https://github.com/agiresearch/AIOS-LSFS)

364、SonicSim: A customizable simulation platform for speech processing in moving sound source scenarios

- Paper: [**点击下载**](https://openreview.net/attachment?id=Hx2ADQLi8M&name=pdf)
- Code: [https://cslikai.cn/SonicSim/](https://cslikai.cn/SonicSim/)

365、Decoupled Graph Energy-based Model for Node Out-of-Distribution Detection on Heterophilic Graphs

- Paper: [**点击下载**](https://openreview.net/attachment?id=NuVBI4wPMm&name=pdf)
- Code: [https://github.com/draym28/DeGEM](https://github.com/draym28/DeGEM)

366、BaB-ND: Long-Horizon Motion Planning with Branch-and-Bound and Neural Dynamics

- Paper: [**点击下载**](https://openreview.net/attachment?id=JXKFPJe0NU&name=pdf)
- Code: [https://robopil.github.io/bab-nd/](https://robopil.github.io/bab-nd/)

367、BigDocs: An Open Dataset for Training Multimodal Models on Document and Code Tasks

- Paper: [**点击下载**](https://openreview.net/attachment?id=b1ivBPLb1n&name=pdf)
- Code: [https://bigdocs.github.io](https://bigdocs.github.io)

368、HERO: Human-Feedback Efficient Reinforcement Learning for Online Diffusion Model Finetuning

- Paper: [**点击下载**](https://openreview.net/attachment?id=yMHe9SRvxk&name=pdf)
- Code: [https://hero-dm.github.io/](https://hero-dm.github.io/)

369、Fully-inductive Node Classification on Arbitrary Graphs

- Paper: [**点击下载**](https://openreview.net/attachment?id=1Qpt43cqhg&name=pdf)
- Code: [https://github.com/DeepGraphLearning/GraphAny](https://github.com/DeepGraphLearning/GraphAny)

370、Multimodal Large Language Models for Inverse Molecular Design with Retrosynthetic Planning

- Paper: [**点击下载**](https://openreview.net/attachment?id=rQ7fz9NO7f&name=pdf)
- Code: [https://github.com/liugangcode/Llamole](https://github.com/liugangcode/Llamole)

371、Symbolic regression via MDLformer-guided search: from minimizing prediction error to minimizing description length

- Paper: [**点击下载**](https://openreview.net/attachment?id=ljAS7cPAU0&name=pdf)
- Code: [https://github.com/tsinghua-fib-lab/SR4MDL](https://github.com/tsinghua-fib-lab/SR4MDL)

372、Detecting Backdoor Samples in Contrastive Language Image Pretraining

- Paper: [**点击下载**](https://openreview.net/attachment?id=KmQEsIfhr9&name=pdf)
- Code: [https://github.com/HanxunH/Detect-CLIP-Backdoor-Samples](https://github.com/HanxunH/Detect-CLIP-Backdoor-Samples)

373、Guided Score identity Distillation for Data-Free One-Step Text-to-Image Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=HMVDiaWMwM&name=pdf)
- Code: [https://github.com/mingyuanzhou/SiD-LSG](https://github.com/mingyuanzhou/SiD-LSG)

374、Competing Large Language Models in Multi-Agent Gaming Environments

- Paper: [**点击下载**](https://openreview.net/attachment?id=DI4gW8viB6&name=pdf)
- Code: [https://github.com/CUHK-ARISE/GAMABench](https://github.com/CUHK-ARISE/GAMABench)

375、Reinforcement learning with combinatorial actions for coupled restless bandits

- Paper: [**点击下载**](https://openreview.net/attachment?id=DhH3LbA6F6&name=pdf)
- Code: [https://github.com/lily-x/combinatorial-rmab](https://github.com/lily-x/combinatorial-rmab)

376、OGBench: Benchmarking Offline Goal-Conditioned RL

- Paper: [**点击下载**](https://openreview.net/attachment?id=M992mjgKzI&name=pdf)
- Code: [https://seohong.me/projects/ogbench](https://seohong.me/projects/ogbench)

377、CREMA: Generalizable and Efficient Video-Language Reasoning via Multimodal Modular Fusion

- Paper: [**点击下载**](https://openreview.net/attachment?id=3UaOlzDEt2&name=pdf)
- Code: [https://CREMA-VideoLLM.github.io/](https://CREMA-VideoLLM.github.io/)

378、Aligned LLMs Are Not Aligned Browser Agents

- Paper: [**点击下载**](https://openreview.net/attachment?id=NsFZZU9gvk&name=pdf)
- Code: [https://huggingface.co/datasets/ScaleAI/BrowserART](https://huggingface.co/datasets/ScaleAI/BrowserART)

379、Edge Prompt Tuning for Graph Neural Networks

- Paper: [**点击下载**](https://openreview.net/attachment?id=92vMaHotTM&name=pdf)
- Code: [https://github.com/xbfu/EdgePrompt](https://github.com/xbfu/EdgePrompt)

380、Optimized Multi-Token Joint Decoding With Auxiliary Model for LLM Inference

- Paper: [**点击下载**](https://openreview.net/attachment?id=ZHhBawo3k5&name=pdf)
- Code: [https://github.com/ZongyueQin/MTAD](https://github.com/ZongyueQin/MTAD)

381、Algorithmic Stability Based Generalization Bounds for Adversarial Training

- Paper: [**点击下载**](https://openreview.net/attachment?id=2GwMazl9ND&name=pdf)
- Code: [https://github.com/rzTian/AT-Stability](https://github.com/rzTian/AT-Stability)

382、Learning a Neural Solver for Parametric PDEs to Enhance Physics-Informed Methods

- Paper: [**点击下载**](https://openreview.net/attachment?id=jqVj8vCQsT&name=pdf)
- Code: [https://github.com/2ailesB/neural-parametric-solver](https://github.com/2ailesB/neural-parametric-solver)

383、MS-Diffusion: Multi-subject Zero-shot Image Personalization with Layout Guidance

- Paper: [**点击下载**](https://openreview.net/attachment?id=PJqP0wyQek&name=pdf)
- Code: [https://MS-Diffusion.github.io](https://MS-Diffusion.github.io)

384、AdaManip: Adaptive Articulated Object Manipulation Environments and Policy Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=Luss2sa0vc&name=pdf)
- Code: [https://adamanip.github.io](https://adamanip.github.io)

385、Zero-shot Model-based Reinforcement Learning using Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=uZFXpPrwSh&name=pdf)
- Code: [https://github.com/abenechehab/dicl](https://github.com/abenechehab/dicl)

386、Interpreting the Second-Order Effects of Neurons in CLIP

- Paper: [**点击下载**](https://openreview.net/attachment?id=GPDcvoFGOL&name=pdf)
- Code: [https://yossigandelsman.github.io/clip_neurons/](https://yossigandelsman.github.io/clip_neurons/)

387、GOAL: A Generalist Combinatorial Optimization Agent Learner

- Paper: [**点击下载**](https://openreview.net/attachment?id=z2z9suDRjw&name=pdf)
- Code: [https://github.com/naver/goal-co](https://github.com/naver/goal-co)

388、Cross the Gap:  Exposing the Intra-modal Misalignment in CLIP via Modality Inversion

- Paper: [**点击下载**](https://openreview.net/attachment?id=VVVfuIcmKR&name=pdf)
- Code: [https://github.com/miccunifi/Cross-the-Gap](https://github.com/miccunifi/Cross-the-Gap)

389、Reading Your Heart: Learning ECG Words and Sentences via Pre-training ECG Language Model

- Paper: [**点击下载**](https://openreview.net/attachment?id=6Hz1Ko087B&name=pdf)
- Code: [https://github.com/PKUDigitalHealth/HeartLang](https://github.com/PKUDigitalHealth/HeartLang)

390、Dynamic-LLaVA: Efficient Multimodal Large Language Models via Dynamic Vision-language Context Sparsification

- Paper: [**点击下载**](https://openreview.net/attachment?id=hzVpZDrW73&name=pdf)
- Code: [https://github.com/Osilly/dynamic_llava](https://github.com/Osilly/dynamic_llava)

391、Improved Training Technique for Latent Consistency Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=PQjZes6vFV&name=pdf)
- Code: [https://github.com/quandao10/sLCT/](https://github.com/quandao10/sLCT/)

392、Dynamic Negative Guidance of Diffusion Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=6p74UyAdLa&name=pdf)
- Code: [https://github.com/FelixKoulischer/Dynamic-Negative-Guidance.git](https://github.com/FelixKoulischer/Dynamic-Negative-Guidance.git)

393、Adversarial Training for Defense Against Label Poisoning Attacks

- Paper: [**点击下载**](https://openreview.net/attachment?id=UlpkHciYQP&name=pdf)
- Code: [https://github.com/melisilaydabal/floral](https://github.com/melisilaydabal/floral)

394、Trajectory attention for fine-grained video motion control

- Paper: [**点击下载**](https://openreview.net/attachment?id=2z1HT5lw5M&name=pdf)
- Code: [https://xizaoqu.github.io/trajattn/](https://xizaoqu.github.io/trajattn/)

395、Diff-Prompt: Diffusion-driven Prompt Generator with Mask Supervision

- Paper: [**点击下载**](https://openreview.net/attachment?id=LfghnrSJNg&name=pdf)
- Code: [https://github.com/Kelvin-ywc/diff-prompt](https://github.com/Kelvin-ywc/diff-prompt)

396、LDAdam: Adaptive Optimization from Low-Dimensional Gradient Statistics

- Paper: [**点击下载**](https://openreview.net/attachment?id=Zkp1GuHerF&name=pdf)
- Code: [https://github.com/IST-DASLab/LDAdam](https://github.com/IST-DASLab/LDAdam)

397、Towards Robust Multimodal Open-set Test-time Adaptation via Adaptive Entropy-aware Optimization

- Paper: [**点击下载**](https://openreview.net/attachment?id=hj323oR3rw&name=pdf)
- Code: [https://github.com/donghao51/AEO](https://github.com/donghao51/AEO)

398、HALL-E: Hierarchical Neural Codec Language Model for Minute-Long Zero-Shot Text-to-Speech Synthesis

- Paper: [**点击下载**](https://openreview.net/attachment?id=868masI331&name=pdf)
- Code: [https://yutonishimura-v2.github.io/HALL-E_DEMO](https://yutonishimura-v2.github.io/HALL-E_DEMO)

399、A General Framework for Producing Interpretable Semantic Text Embeddings

- Paper: [**点击下载**](https://openreview.net/attachment?id=23uY3FpQxc&name=pdf)
- Code: [https://github.com/dukesun99/CQG-MBQA](https://github.com/dukesun99/CQG-MBQA)

400、Multimodal Situational Safety

- Paper: [**点击下载**](https://openreview.net/attachment?id=I9bEi6LNgt&name=pdf)
- Code: [https://mssbench.github.io/](https://mssbench.github.io/)

401、Visual-O1: Understanding Ambiguous Instructions via Multi-modal Multi-turn Chain-of-thoughts Reasoning

- Paper: [**点击下载**](https://openreview.net/attachment?id=v9CDpLpjiE&name=pdf)
- Code: [https://github.com/kodenii/Visual-O1](https://github.com/kodenii/Visual-O1)

402、Trusted Multi-View Classification via Evolutionary Multi-View Fusion

- Paper: [**点击下载**](https://openreview.net/attachment?id=M3kBtqpys5&name=pdf)
- Code: [https://github.com/fupinhan123/TEF](https://github.com/fupinhan123/TEF)

403、HQGS: High-Quality Novel View Synthesis with Gaussian Splatting in Degraded Scenes

- Paper: [**点击下载**](https://openreview.net/attachment?id=25Zlvl7JxW&name=pdf)
- Code: [https://github.com/linxin0/HQGS](https://github.com/linxin0/HQGS)

404、REFINE: Inversion-Free Backdoor Defense via Model Reprogramming

- Paper: [**点击下载**](https://openreview.net/attachment?id=4IYdCws9fc&name=pdf)
- Code: [https://github.com/WhitolfChen/REFINE](https://github.com/WhitolfChen/REFINE)

405、Tight Clusters Make Specialized Experts

- Paper: [**点击下载**](https://openreview.net/attachment?id=Pu3c0209cx&name=pdf)
- Code: [https://github.com/stefvk/ACMoE](https://github.com/stefvk/ACMoE)

406、Real-Time Video Generation with Pyramid Attention Broadcast

- Paper: [**点击下载**](https://openreview.net/attachment?id=hDBrQ4DApF&name=pdf)
- Code: [https://github.com/NUS-HPC-AI-Lab/OpenDiT](https://github.com/NUS-HPC-AI-Lab/OpenDiT)

407、Large Convolutional Model Tuning via Filter Subspace

- Paper: [**点击下载**](https://openreview.net/attachment?id=E5YmIBvOqV&name=pdf)
- Code: [https://github.com/weichennone/convnet_finetune](https://github.com/weichennone/convnet_finetune)

408、Spectro-Riemannian Graph Neural Networks

- Paper: [**点击下载**](https://openreview.net/attachment?id=2MLvV7fvAz&name=pdf)
- Code: [https://github.com/amazon-science/cusp](https://github.com/amazon-science/cusp)

409、Towards Out-of-Modal Generalization without Instance-level Modal Correspondence

- Paper: [**点击下载**](https://openreview.net/attachment?id=LuVulfPgZN&name=pdf)
- Code: [https://github.com/tmllab/2025_ICLR_COX](https://github.com/tmllab/2025_ICLR_COX)

410、Subtask-Aware Visual Reward Learning from Segmented Demonstrations

- Paper: [**点击下载**](https://openreview.net/attachment?id=mqKVe6F3Up&name=pdf)
- Code: [https://changyeon.site/reds](https://changyeon.site/reds)

411、TopoDiffusionNet: A Topology-aware Diffusion Model

- Paper: [**点击下载**](https://openreview.net/attachment?id=ZK1LoTo10R&name=pdf)
- Code: [https://github.com/Saumya-Gupta-26/TopoDiffusionNet](https://github.com/Saumya-Gupta-26/TopoDiffusionNet)

412、3DGS-Drag: Dragging Gaussians for Intuitive Point-Based 3D Editing

- Paper: [**点击下载**](https://openreview.net/attachment?id=7JUrBLDjCq&name=pdf)
- Code: [https://github.com/Dongjiahua/3DGS-Drag](https://github.com/Dongjiahua/3DGS-Drag)

413、Learning Generalizable Skills from Offline Multi-Task Data for Multi-Agent Cooperation

- Paper: [**点击下载**](https://openreview.net/attachment?id=HR1ujVR0ig&name=pdf)
- Code: [https://github.com/mooricAnna/HiSSD](https://github.com/mooricAnna/HiSSD)

414、What Matters When Repurposing Diffusion Models for General Dense Perception Tasks?

- Paper: [**点击下载**](https://openreview.net/attachment?id=BgYbk6ZmeX&name=pdf)
- Code: [https://github.com/aim-uofa/GenPercept](https://github.com/aim-uofa/GenPercept)

415、Adversarially Robust Out-of-Distribution Detection Using Lyapunov-Stabilized Embeddings

- Paper: [**点击下载**](https://openreview.net/attachment?id=GrDne4055L&name=pdf)
- Code: [https://github.com/AdaptiveMotorControlLab/AROS](https://github.com/AdaptiveMotorControlLab/AROS)

416、Simulating Human-like Daily Activities with Desire-driven Autonomy

- Paper: [**点击下载**](https://openreview.net/attachment?id=3ms8EQY7f8&name=pdf)
- Code: [https://sites.google.com/view/desire-driven-autonomy](https://sites.google.com/view/desire-driven-autonomy)

417、CtrLoRA: An Extensible and Efficient Framework for Controllable Image Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=3Gga05Jdmj&name=pdf)
- Code: [https://github.com/xyfJASON/ctrlora](https://github.com/xyfJASON/ctrlora)

418、ParaSolver: A Hierarchical Parallel Integral Solver for Diffusion Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=2JihLwirxO&name=pdf)
- Code: [https://github.com/Jianrong-Lu/ParaSolver.git](https://github.com/Jianrong-Lu/ParaSolver.git)

419、Mining your own secrets: Diffusion Classifier Scores for Continual Personalization of Text-to-Image Diffusion Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=hUdLs6TqZL&name=pdf)
- Code: [https://srvcodes.github.io/continual_personalization/](https://srvcodes.github.io/continual_personalization/)

420、RAG-DDR: Optimizing Retrieval-Augmented Generation Using Differentiable Data Rewards

- Paper: [**点击下载**](https://openreview.net/attachment?id=Pnktu2PBXD&name=pdf)
- Code: [https://github.com/OpenMatch/RAG-DDR](https://github.com/OpenMatch/RAG-DDR)

421、GeoX: Geometric Problem Solving Through Unified Formalized Vision-Language Pre-training

- Paper: [**点击下载**](https://openreview.net/attachment?id=6RiBl5sCDF&name=pdf)
- Code: [https://github.com/Alpha-Innovator/GeoX](https://github.com/Alpha-Innovator/GeoX)

422、Two Sparse Matrices are Better than One: Sparsifying Neural Networks with Double Sparse Factorization

- Paper: [**点击下载**](https://openreview.net/attachment?id=DwiwOcK1B7&name=pdf)
- Code: [https://github.com/usamec/double_sparse](https://github.com/usamec/double_sparse)

423、Motion-Agent: A Conversational Framework for Human Motion Generation with LLMs

- Paper: [**点击下载**](https://openreview.net/attachment?id=AvOhBgsE5R&name=pdf)
- Code: [https://knoxzhao.github.io/Motion-Agent](https://knoxzhao.github.io/Motion-Agent)

424、AstroCompress: A benchmark dataset for multi-purpose compression of astronomical data

- Paper: [**点击下载**](https://openreview.net/attachment?id=kQCHCkNk7s&name=pdf)
- Code: [https://huggingface.co/AstroCompress](https://huggingface.co/AstroCompress)

425、DCT-CryptoNets: Scaling Private Inference in the Frequency Domain

- Paper: [**点击下载**](https://openreview.net/attachment?id=lPJUQsSIxm&name=pdf)
- Code: [https://github.com/ar-roy/dct-cryptonets](https://github.com/ar-roy/dct-cryptonets)

426、Unbounded: A Generative Infinite Game of Character Life Simulation

- Paper: [**点击下载**](https://openreview.net/attachment?id=uy31tqVuNo&name=pdf)
- Code: [https://infinite-generative-game.github.io/](https://infinite-generative-game.github.io/)

427、Metric-Driven Attributions for Vision Transformers

- Paper: [**点击下载**](https://openreview.net/attachment?id=rGP2jbWt0l&name=pdf)
- Code: [https://github.com/chasewalker26/MDA-Metric-Driven-Attributions-for-ViT](https://github.com/chasewalker26/MDA-Metric-Driven-Attributions-for-ViT)

428、Diffusion Models Are Real-Time Game Engines

- Paper: [**点击下载**](https://openreview.net/attachment?id=P8pqeEkn1H&name=pdf)
- Code: [https://www.reddit.com/r/itrunsdoom/](https://www.reddit.com/r/itrunsdoom/)

429、Going Beyond Feature Similarity: Effective Dataset distillation based on Class-aware Conditional Mutual Information

- Paper: [**点击下载**](https://openreview.net/attachment?id=0no1Wp2R2j&name=pdf)
- Code: [https://github.com/ndhg1213/CMIDD](https://github.com/ndhg1213/CMIDD)

430、GravMAD: Grounded Spatial Value Maps Guided Action Diffusion for Generalized 3D Manipulation

- Paper: [**点击下载**](https://openreview.net/attachment?id=qPzYF2EpXb&name=pdf)
- Code: [yangtaochen@smali.nju.edu.cn, {chenzx, huojing, gaoy} @nju.edu.cn,](yangtaochen@smali.nju.edu.cn, {chenzx, huojing, gaoy} @nju.edu.cn,)

431、DriveTransformer: Unified Transformer for Scalable End-to-End Autonomous Driving

- Paper: [**点击下载**](https://openreview.net/attachment?id=M42KR4W9P5&name=pdf)
- Code: [https://github.com/Thinklab-SJTU/DriveTransformer/](https://github.com/Thinklab-SJTU/DriveTransformer/)

432、N-ForGOT: Towards Not-forgetting and Generalization of Open Temporal Graph Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=rLlDt2FQvz&name=pdf)
- Code: [https://github.com/Liane-Wang/N-ForGOT](https://github.com/Liane-Wang/N-ForGOT)

433、Rethinking Light Decoder-based Solvers for Vehicle Routing Problems

- Paper: [**点击下载**](https://openreview.net/attachment?id=4pRwkYpa2u&name=pdf)
- Code: [https://github.com/ziweileonhuang/reld-nco](https://github.com/ziweileonhuang/reld-nco)

434、Rethinking Diffusion Posterior Sampling: From Conditional Score Estimator to Maximizing a Posterior

- Paper: [**点击下载**](https://openreview.net/attachment?id=GcvLoqOoXL&name=pdf)
- Code: [https://github.com/tongdaxu/Rethinking-Diffusion-Posterior-Sampling-From-Conditional-Score-Estimator-to-Maximizing-a-Posterior](https://github.com/tongdaxu/Rethinking-Diffusion-Posterior-Sampling-From-Conditional-Score-Estimator-to-Maximizing-a-Posterior)

435、Temporal Flexibility in Spiking Neural Networks: Towards Generalization Across Time Steps and Deployment Friendliness

- Paper: [**点击下载**](https://openreview.net/attachment?id=9HsfTgflT7&name=pdf)
- Code: [https://github.com/brain-intelligence-lab/temporal_flexibility_in_SNN](https://github.com/brain-intelligence-lab/temporal_flexibility_in_SNN)

436、Slot-Guided Adaptation of Pre-trained Diffusion Models for Object-Centric Learning and Compositional Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=kZvor5aaz7&name=pdf)
- Code: [https://kaanakan.github.io/SlotAdapt/](https://kaanakan.github.io/SlotAdapt/)

437、SpinQuant: LLM Quantization with Learned Rotations

- Paper: [**点击下载**](https://openreview.net/attachment?id=ogO6DGE6FZ&name=pdf)
- Code: [https://github.com/facebookresearch/SpinQuant](https://github.com/facebookresearch/SpinQuant)

438、On the Optimization Landscape of Low Rank Adaptation Methods for Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=pxclAomHat&name=pdf)
- Code: [https://github.com/liuxhym/GaRare.git](https://github.com/liuxhym/GaRare.git)

439、Are Transformers Able to Reason by Connecting Separated Knowledge in Training Data?

- Paper: [**点击下载**](https://openreview.net/attachment?id=1Xg4JPPxJ0&name=pdf)
- Code: [https://github.com/YYT-t/Fragmented-at-Training-Chained-at-Testing](https://github.com/YYT-t/Fragmented-at-Training-Chained-at-Testing)

440、A Percolation Model of Emergence: Analyzing Transformers Trained on a Formal Language

- Paper: [**点击下载**](https://openreview.net/attachment?id=0pLCDJVVRD&name=pdf)
- Code: [https://github.com/EkdeepSLubana/ConceptPercolation](https://github.com/EkdeepSLubana/ConceptPercolation)

441、Adaptive Retention & Correction: Test-Time Training for Continual Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=9bLdbp46Q1&name=pdf)
- Code: [https://github.com/HaoranChen/Adaptive-Retention-and-Correction-for-Continual-Learning](https://github.com/HaoranChen/Adaptive-Retention-and-Correction-for-Continual-Learning)

442、Do WGANs succeed because they minimize the Wasserstein Distance? Lessons from Discrete Generators

- Paper: [**点击下载**](https://openreview.net/attachment?id=7YXaOvunqo&name=pdf)
- Code: [https://github.com/ariel415el/DiscreteGANs.git](https://github.com/ariel415el/DiscreteGANs.git)

443、Immunogenicity Prediction with Dual Attention Enables Vaccine Target Selection

- Paper: [**点击下载**](https://openreview.net/attachment?id=hWmwL9gizZ&name=pdf)
- Code: [https://github.com/songleee/VenusVaccine](https://github.com/songleee/VenusVaccine)

444、Correlating instruction-tuning (in multimodal models) with vision-language processing (in the brain)

- Paper: [**点击下载**](https://openreview.net/attachment?id=xkgfLXZ4e0&name=pdf)
- Code: [https://github.com/subbareddy248/mllm_instruction_brain](https://github.com/subbareddy248/mllm_instruction_brain)

445、WavTokenizer: an Efficient Acoustic Discrete Codec Tokenizer for Audio Language Modeling

- Paper: [**点击下载**](https://openreview.net/attachment?id=yBlVlS2Fd9&name=pdf)
- Code: [https://github.com/jishengpeng/WavTokenizer](https://github.com/jishengpeng/WavTokenizer)

446、Bad-PFL: Exploiting Backdoor Attacks against Personalized Federated Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=79nO2DPjVX&name=pdf)
- Code: [https://github.com/fmy266/Bad-PFL](https://github.com/fmy266/Bad-PFL)

447、Enhancing Multilingual Reasoning in LLMs: Insights from Cross-Linguistic Correlations and Optimal Data Proportions

- Paper: [**点击下载**](https://openreview.net/attachment?id=S6cBH99BhB&name=pdf)
- Code: [https://github.com/DeepShareAI/HighMath](https://github.com/DeepShareAI/HighMath)

448、PhysPDE: Rethinking PDE Discovery and a Physical HYpothesis Selection Benchmark

- Paper: [**点击下载**](https://openreview.net/attachment?id=G3CpBCQwNh&name=pdf)
- Code: [https://github.com/FengMingquan-sjtu/PhysPDE](https://github.com/FengMingquan-sjtu/PhysPDE)

449、Intervening Anchor Token: Decoding Strategy in Alleviating Hallucinations for MLLMs

- Paper: [**点击下载**](https://openreview.net/attachment?id=zGb4WgCW5i&name=pdf)
- Code: [https://github.com/Everlyn-Labs/ANTRP](https://github.com/Everlyn-Labs/ANTRP)

450、Mitigating Modality Prior-Induced Hallucinations in Multimodal Large Language Models via Deciphering Attention Causality

- Paper: [**点击下载**](https://openreview.net/attachment?id=AV7OXVlAyi&name=pdf)
- Code: [https://github.com/The-Martyr/CausalMM](https://github.com/The-Martyr/CausalMM)

451、Improving Data Efficiency via Curating LLM-Driven Rating Systems

- Paper: [**点击下载**](https://openreview.net/attachment?id=DKkQtRMowq&name=pdf)
- Code: [https://github.com/UCSC-REAL/DS2](https://github.com/UCSC-REAL/DS2)

452、VL-ICL Bench: The Devil in the Details of Multimodal In-Context Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=cpGPPLLYYx&name=pdf)
- Code: [https://ys-zong.github.io/VL-ICL/](https://ys-zong.github.io/VL-ICL/)

453、ECHOPulse: ECG Controlled Echocardio-gram Video Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=i2r7LDjba3&name=pdf)
- Code: [https://github.com/levyisthebest/ECHOPulse_Prelease](https://github.com/levyisthebest/ECHOPulse_Prelease)

454、A Single Goal is All You Need: Skills and Exploration Emerge from Contrastive RL without Rewards, Demonstrations, or Subgoals

- Paper: [**点击下载**](https://openreview.net/attachment?id=xCkgX4Xfu0&name=pdf)
- Code: [https://graliuce.github.io/sgcrl/](https://graliuce.github.io/sgcrl/)

455、Hummingbird: High Fidelity Image Generation via Multimodal Context Alignment

- Paper: [**点击下载**](https://openreview.net/attachment?id=6kPBThI6ZJ&name=pdf)
- Code: [https://roar-ai.github.io/hummingbird](https://roar-ai.github.io/hummingbird)

456、Causal Graph Transformer for Treatment Effect Estimation Under Unknown Interference

- Paper: [**点击下载**](https://openreview.net/attachment?id=foQ4AeEGG7&name=pdf)
- Code: [https://github.com/anpwu/CauGramer](https://github.com/anpwu/CauGramer)

457、Phidias: A Generative Model for Creating 3D  Content from Text, Image, and 3D Conditions with Reference-Augmented  Diffusion

- Paper: [**点击下载**](https://openreview.net/attachment?id=TEkoMEjf7E&name=pdf)
- Code: [https://RAG-3D.github.io/](https://RAG-3D.github.io/)

458、Multiview Equivariance Improves 3D Correspondence Understanding with Minimal Feature Finetuning

- Paper: [**点击下载**](https://openreview.net/attachment?id=CNO4rbSV6v&name=pdf)
- Code: [https://github.com/qq456cvb/3DCorrEnhance](https://github.com/qq456cvb/3DCorrEnhance)

459、Learning Gain Map for Inverse Tone Mapping

- Paper: [**点击下载**](https://openreview.net/attachment?id=GtHRhpgpzB&name=pdf)
- Code: [https://github.com/qtlark/GMNet](https://github.com/qtlark/GMNet)

460、Latent Radiance Fields with 3D-aware 2D Representations

- Paper: [**点击下载**](https://openreview.net/attachment?id=vL9t9tpKli&name=pdf)
- Code: [https://latent-radiance-field.github.io/LRF](https://latent-radiance-field.github.io/LRF)

461、Speculative Knowledge Distillation: Bridging the Teacher-Student Gap Through Interleaved Sampling

- Paper: [**点击下载**](https://openreview.net/attachment?id=EgJhwYR2tB&name=pdf)
- Code: [https://github.com/xu1998hz/skd](https://github.com/xu1998hz/skd)

462、SMI-Editor: Edit-based SMILES Language Model with Fragment-level Supervision

- Paper: [**点击下载**](https://openreview.net/attachment?id=M29nUGozPa&name=pdf)
- Code: [https://github.com/zhengkangjie/smi-editor](https://github.com/zhengkangjie/smi-editor)

463、Long-Sequence Recommendation Models Need Decoupled Embeddings

- Paper: [**点击下载**](https://openreview.net/attachment?id=jkpGIxSsUD&name=pdf)
- Code: [https://github.com/thuml/DARE](https://github.com/thuml/DARE)

464、TIPS: Text-Image Pretraining with Spatial awareness

- Paper: [**点击下载**](https://openreview.net/attachment?id=DaA0wAcTY7&name=pdf)
- Code: [https://github.com/google-deepmind/tips](https://github.com/google-deepmind/tips)

465、OMNI-EPIC: Open-endedness via Models of human Notions of Interestingness with Environments Programmed in Code

- Paper: [**点击下载**](https://openreview.net/attachment?id=Y1XkzMJpPd&name=pdf)
- Code: [https://dub.sh/omniepic](https://dub.sh/omniepic)

466、Cut the Crap: An Economical Communication Pipeline for LLM-based Multi-Agent Systems

- Paper: [**点击下载**](https://openreview.net/attachment?id=LkzuPorQ5L&name=pdf)
- Code: [https://anonymous.4open.science/r/AgentPrune-F6A5](https://anonymous.4open.science/r/AgentPrune-F6A5)

467、Halton Scheduler for Masked Generative Image Transformer

- Paper: [**点击下载**](https://openreview.net/attachment?id=RDVrlWAb7K&name=pdf)
- Code: [https://github.com/valeoai/Halton-MaskGIT](https://github.com/valeoai/Halton-MaskGIT)

468、MVTokenFlow: High-quality 4D Content Generation using Multiview Token Flow

- Paper: [**点击下载**](https://openreview.net/attachment?id=zu7cBTPsDb&name=pdf)
- Code: [https://soolab.github.io/MVTokenFlow](https://soolab.github.io/MVTokenFlow)

469、Black Sheep in the Herd: Playing with Spuriously Correlated Attributes for Vision-Language Recognition

- Paper: [**点击下载**](https://openreview.net/attachment?id=g1fkhbhHjL&name=pdf)
- Code: [https://github.com/Liam-Tian/sas](https://github.com/Liam-Tian/sas)

470、Framer: Interactive Frame Interpolation

- Paper: [**点击下载**](https://openreview.net/attachment?id=Lp40Z40N07&name=pdf)
- Code: [https://github.com/aim-uofa/Framer](https://github.com/aim-uofa/Framer)

471、GS-CPR: Efficient Camera Pose Refinement via 3D Gaussian Splatting

- Paper: [**点击下载**](https://openreview.net/attachment?id=mP7uV59iJM&name=pdf)
- Code: [https://xrim-lab.github.io/GS-CPR/](https://xrim-lab.github.io/GS-CPR/)

472、BrainUICL: An Unsupervised Individual Continual Learning Framework for EEG Applications

- Paper: [**点击下载**](https://openreview.net/attachment?id=6jjAYmppGQ&name=pdf)
- Code: [https://github.com/xiaobaben/BrainUICL](https://github.com/xiaobaben/BrainUICL)

473、Is In-Context Learning Sufficient for Instruction Following in LLMs?

- Paper: [**点击下载**](https://openreview.net/attachment?id=STEEDDv3zI&name=pdf)
- Code: [https://github.com/tml-epfl/icl-alignment](https://github.com/tml-epfl/icl-alignment)

474、Towards Effective Evaluations and Comparisons for LLM Unlearning Methods

- Paper: [**点击下载**](https://openreview.net/attachment?id=wUtCieKuQU&name=pdf)
- Code: [https://github.com/tmlr-group/Unlearning-with-Control](https://github.com/tmlr-group/Unlearning-with-Control)

475、Learning Evolving Tools for Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=wtrDLMFU9v&name=pdf)
- Code: [https://github.com/Chen-GX/ToolEVO](https://github.com/Chen-GX/ToolEVO)

476、Distilling Reinforcement Learning Algorithms for In-Context Model-Based Planning

- Paper: [**点击下载**](https://openreview.net/attachment?id=BfUugGfBE5&name=pdf)
- Code: [https://github.com/jaehyeon-son/dicp](https://github.com/jaehyeon-son/dicp)

477、OVTR: End-to-End Open-Vocabulary Multiple Object Tracking with Transformer

- Paper: [**点击下载**](https://openreview.net/attachment?id=GDS5eN65QY&name=pdf)
- Code: [https://github.com/jinyanglii/OVTR](https://github.com/jinyanglii/OVTR)

478、BOFormer: Learning to Solve Multi-Objective Bayesian Optimization via Non-Markovian RL

- Paper: [**点击下载**](https://openreview.net/attachment?id=UnCKU8pZVe&name=pdf)
- Code: [https://hungyuheng.github.io/BOFormer/](https://hungyuheng.github.io/BOFormer/)

479、RDT-1B: a Diffusion Foundation Model for Bimanual Manipulation

- Paper: [**点击下载**](https://openreview.net/attachment?id=yAzN4tz7oI&name=pdf)
- Code: [https://rdt-robotics.github.io/rdt-robotics/](https://rdt-robotics.github.io/rdt-robotics/)

480、VILA-U: a Unified Foundation Model Integrating Visual Understanding and Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=02haSpO453&name=pdf)
- Code: [https://hanlab.mit.edu/projects/vila-u](https://hanlab.mit.edu/projects/vila-u)

481、GS-LiDAR: Generating Realistic LiDAR Point Clouds with Panoramic Gaussian Splatting

- Paper: [**点击下载**](https://openreview.net/attachment?id=RMaRBE9s2H&name=pdf)
- Code: [https://github.com/fudan-zvg/GS-LiDAR](https://github.com/fudan-zvg/GS-LiDAR)

482、Chain-of-Focus Prompting: Leveraging Sequential Visual Cues to Prompt Large Autoregressive Vision Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=noidywkBba&name=pdf)
- Code: [https://github.com/tmllab/2025_ICLR_COF](https://github.com/tmllab/2025_ICLR_COF)

483、PAL: Sample-Efficient Personalized Reward Modeling for Pluralistic Alignment

- Paper: [**点击下载**](https://openreview.net/attachment?id=1kFDrYCuSu&name=pdf)
- Code: [https://github.com/RamyaLab/pluralistic-alignment](https://github.com/RamyaLab/pluralistic-alignment)

484、Dynamic Diffusion Transformer

- Paper: [**点击下载**](https://openreview.net/attachment?id=taHwqSrbrb&name=pdf)
- Code: [https://github.com/NUS-HPC-AI-Lab/Dynamic-Diffusion-Transformer](https://github.com/NUS-HPC-AI-Lab/Dynamic-Diffusion-Transformer)

485、McEval: Massively Multilingual Code Evaluation

- Paper: [**点击下载**](https://openreview.net/attachment?id=UunCPtPOlZ&name=pdf)
- Code: [https://github.com/MCEVAL/McEval](https://github.com/MCEVAL/McEval)

486、SEAL: Safety-enhanced Aligned LLM Fine-tuning via Bilevel Data Selection

- Paper: [**点击下载**](https://openreview.net/attachment?id=VHguhvcoM5&name=pdf)
- Code: [https://github.com/hanshen95/SEAL](https://github.com/hanshen95/SEAL)

487、Bidirectional Decoding: Improving Action Chunking via Guided Test-Time Sampling

- Paper: [**点击下载**](https://openreview.net/attachment?id=qZmn2hkuzw&name=pdf)
- Code: [https://bid-robot.github.io](https://bid-robot.github.io)

488、Improving Language Model Distillation through Hidden State Matching

- Paper: [**点击下载**](https://openreview.net/attachment?id=IcVSKhVpKu&name=pdf)
- Code: [https://github.com/Sayan21/ICLR25-CKA](https://github.com/Sayan21/ICLR25-CKA)

489、Navigating Neural Space: Revisiting Concept Activation Vectors to Overcome Directional Divergence

- Paper: [**点击下载**](https://openreview.net/attachment?id=Q95MaWfF4e&name=pdf)
- Code: [https://github.com/frederikpahde/pattern-cav](https://github.com/frederikpahde/pattern-cav)

490、Medium-Difficulty Samples Constitute Smoothed Decision Boundary for Knowledge Distillation on Pruned Datasets

- Paper: [**点击下载**](https://openreview.net/attachment?id=Rz4UkJziFe&name=pdf)
- Code: [https://github.com/chenyd7/MDSLR](https://github.com/chenyd7/MDSLR)

491、Revisiting Convolution Architecture in the Realm of DNA Foundation Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=B07dLVWLyD&name=pdf)
- Code: [https://github.com/aim-uofa/ConvNova](https://github.com/aim-uofa/ConvNova)

492、Episodic Novelty Through Temporal Distance

- Paper: [**点击下载**](https://openreview.net/attachment?id=I7DeajDEx7&name=pdf)
- Code: [https://github.com/Jackory/ETD](https://github.com/Jackory/ETD)

493、Scaling Autonomous Agents via Automatic Reward Modeling And Planning

- Paper: [**点击下载**](https://openreview.net/attachment?id=womU9cEwcO&name=pdf)
- Code: [https://armap-agent.github.io](https://armap-agent.github.io)

494、FlickerFusion: Intra-trajectory Domain Generalizing Multi-agent Reinforcement Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=MRYyOaNxh3&name=pdf)
- Code: [flickerfusion305.github.io](flickerfusion305.github.io)

495、Gradient-Free Generation for Hard-Constrained Systems

- Paper: [**点击下载**](https://openreview.net/attachment?id=teE4pl9ftK&name=pdf)
- Code: [https://github.com/amazon-science/ECI-sampling](https://github.com/amazon-science/ECI-sampling)

496、M^3PC: Test-time Model Predictive Control using Pretrained Masked Trajectory Model

- Paper: [**点击下载**](https://openreview.net/attachment?id=inOwd7hZC1&name=pdf)
- Code: [https://github.com/wkh923/m3pc](https://github.com/wkh923/m3pc)

497、Dynamical Diffusion: Learning Temporal Dynamics with Diffusion Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=c5JZEPyFUE&name=pdf)
- Code: [https://github.com/thuml/dynamical-diffusion](https://github.com/thuml/dynamical-diffusion)

498、ProtoSnap: Prototype Alignment For Cuneiform Signs

- Paper: [**点击下载**](https://openreview.net/attachment?id=XHTirKsQV6&name=pdf)
- Code: [https://tau-vailab.github.io/ProtoSnap/](https://tau-vailab.github.io/ProtoSnap/)

499、DiffSplat: Repurposing Image Diffusion Models for Scalable Gaussian Splat Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=eajZpoQkGK&name=pdf)
- Code: [https://chenguolin.github.io/projects/DiffSplat](https://chenguolin.github.io/projects/DiffSplat)

500、Learning Spatial-Semantic Features for Robust Video Object Segmentation

- Paper: [**点击下载**](https://openreview.net/attachment?id=EM93t94zEi&name=pdf)
- Code: [https://github.com/yahooo-m/S3](https://github.com/yahooo-m/S3)

501、Toward Generalizing Visual Brain Decoding to Unseen Subjects

- Paper: [**点击下载**](https://openreview.net/attachment?id=At9JmGF3xy&name=pdf)
- Code: [https://github.com/Xiangtaokong/TGBD](https://github.com/Xiangtaokong/TGBD)

502、On Discriminative Probabilistic Modeling for Self-Supervised Representation Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=s15HrqCqbr&name=pdf)
- Code: [https://github.com/bokun-wang/NUCLR](https://github.com/bokun-wang/NUCLR)

503、Understanding and Mitigating Bottlenecks of State Space Models through the Lens of Recency and Over-smoothing

- Paper: [**点击下载**](https://openreview.net/attachment?id=pymXpl4qvi&name=pdf)
- Code: [https://github.com/VITA-Group/SSM-Bottleneck](https://github.com/VITA-Group/SSM-Bottleneck)

504、PointOBB-v2: Towards Simpler, Faster, and Stronger Single Point Supervised Oriented Object Detection

- Paper: [**点击下载**](https://openreview.net/attachment?id=R22JPTQYWV&name=pdf)
- Code: [https://github.com/taugeren/PointOBB-v2](https://github.com/taugeren/PointOBB-v2)

505、Domain Guidance: A Simple Transfer Approach for a Pre-trained Diffusion Model

- Paper: [**点击下载**](https://openreview.net/attachment?id=PplM2kDrl3&name=pdf)
- Code: [https://github.com/thuml/DomainGuidance.](https://github.com/thuml/DomainGuidance.)

506、MoDGS: Dynamic Gaussian Splatting from Casually-captured Monocular Videos with Depth Priors

- Paper: [**点击下载**](https://openreview.net/attachment?id=2prShxdLkX&name=pdf)
- Code: [https://MoDGS.github.io](https://MoDGS.github.io)

507、VEDIT: Latent Prediction Architecture For Procedural Video Representation Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=LDAj4UJ4aL&name=pdf)
- Code: [https://github.com/HL-hanlin/vedit](https://github.com/HL-hanlin/vedit)

508、SAGEPhos: Sage Bio-Coupled and Augmented Fusion for Phosphorylation Site Detection

- Paper: [**点击下载**](https://openreview.net/attachment?id=hLwcNSFhC2&name=pdf)
- Code: [https://github.com/ZhangJJ26/SAGEPhos](https://github.com/ZhangJJ26/SAGEPhos)

509、Training-Free Dataset Pruning for Instance Segmentation

- Paper: [**点击下载**](https://openreview.net/attachment?id=rvxWEbTtRY&name=pdf)
- Code: [https://github.com/he-y/dataset-pruning-for-instance-segmentation](https://github.com/he-y/dataset-pruning-for-instance-segmentation)

510、PuzzleFusion++: Auto-agglomerative 3D Fracture Assembly by Denoise and Verify

- Paper: [**点击下载**](https://openreview.net/attachment?id=7E7v5mJnfl&name=pdf)
- Code: [https://puzzlefusion-plusplus.github.io](https://puzzlefusion-plusplus.github.io)

511、ET-SEED: EFFICIENT TRAJECTORY-LEVEL SE(3) EQUIVARIANT DIFFUSION POLICY

- Paper: [**点击下载**](https://openreview.net/attachment?id=OheAR2xrtb&name=pdf)
- Code: [https://et-seed.github.io/](https://et-seed.github.io/)

512、Video-STaR: Self-Training Enables Video Instruction Tuning with Any Supervision

- Paper: [**点击下载**](https://openreview.net/attachment?id=JYV2hrtFSv&name=pdf)
- Code: [https://orrzohar.github.io/projects/video-star/](https://orrzohar.github.io/projects/video-star/)

513、Locality Sensitive Avatars From Video

- Paper: [**点击下载**](https://openreview.net/attachment?id=SVta2eQNt3&name=pdf)
- Code: [https://github.com/ChunjinSong/lsavatar](https://github.com/ChunjinSong/lsavatar)

514、Streaming Video Understanding and Multi-round Interaction with Memory-enhanced Knowledge

- Paper: [**点击下载**](https://openreview.net/attachment?id=JbPb6RieNC&name=pdf)
- Code: [https://github.com/hmxiong/StreamChat](https://github.com/hmxiong/StreamChat)

515、LaMP: Language-Motion Pretraining for Motion Generation, Retrieval, and Captioning

- Paper: [**点击下载**](https://openreview.net/attachment?id=LYawG8YkPa&name=pdf)
- Code: [https://aigc3d.github.io/LaMP/](https://aigc3d.github.io/LaMP/)

516、PostCast: Generalizable Postprocessing for Precipitation Nowcasting via Unsupervised Blurriness Modeling

- Paper: [**点击下载**](https://openreview.net/attachment?id=v2zcCDYMok&name=pdf)
- Code: [https://github.com/jasong-ovo/PostCast](https://github.com/jasong-ovo/PostCast)

517、Circuit Representation Learning with Masked Gate Modeling and Verilog-AIG Alignment

- Paper: [**点击下载**](https://openreview.net/attachment?id=US9k5TXVLZ&name=pdf)
- Code: [https://github.com/wuhy68/MGVGA](https://github.com/wuhy68/MGVGA)

518、AuroraCap: Efficient, Performant Video Detailed Captioning and a New Benchmark

- Paper: [**点击下载**](https://openreview.net/attachment?id=tTDUrseRRU&name=pdf)
- Code: [https://rese1f.github.io/aurora-web/](https://rese1f.github.io/aurora-web/)

519、COME: Test-time Adaption by Conservatively Minimizing Entropy

- Paper: [**点击下载**](https://openreview.net/attachment?id=506BjJ1ziZ&name=pdf)
- Code: [https://github.com/BlueWhaleLab/COME](https://github.com/BlueWhaleLab/COME)

520、Fine-Grained Verifiers: Preference Modeling as Next-token Prediction in Vision-Language Alignment

- Paper: [**点击下载**](https://openreview.net/attachment?id=cJQ1K2fjpD&name=pdf)
- Code: [https://github.com/gzcch/FISAO_ICLR](https://github.com/gzcch/FISAO_ICLR)

521、CoInD: Enabling Logical Compositions in Diffusion Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=cCRlEvjrx4&name=pdf)
- Code: [https://github.com/sachit3022/compositional-generation/](https://github.com/sachit3022/compositional-generation/)

522、Automated Design of Agentic Systems

- Paper: [**点击下载**](https://openreview.net/attachment?id=t9U3LW7JVX&name=pdf)
- Code: [https://github.com/ShengranHu/ADAS](https://github.com/ShengranHu/ADAS)

523、ParetoFlow: Guided Flows in Multi-Objective Optimization

- Paper: [**点击下载**](https://openreview.net/attachment?id=mLyyB4le5u&name=pdf)
- Code: [https://github.com/StevenYuan666/ParetoFlow](https://github.com/StevenYuan666/ParetoFlow)

524、Towards Understanding Why Label Smoothing Degrades Selective Classification and How to Fix It

- Paper: [**点击下载**](https://openreview.net/attachment?id=6oWFn6fY4A&name=pdf)
- Code: [https://ensta-u2is-ai.github.io/Understanding-Label-smoothing-Selective-classification/](https://ensta-u2is-ai.github.io/Understanding-Label-smoothing-Selective-classification/)

525、Meissonic: Revitalizing Masked Generative Transformers for Efficient High-Resolution Text-to-Image Synthesis

- Paper: [**点击下载**](https://openreview.net/attachment?id=GJsuYHhAga&name=pdf)
- Code: [https://huggingface.co/MeissonFlow/Meissonic](https://huggingface.co/MeissonFlow/Meissonic)

526、A Coefficient Makes SVRG Effective

- Paper: [**点击下载**](https://openreview.net/attachment?id=twtTLZnG0B&name=pdf)
- Code: [github.com/davidyyd/alpha-SVRG](github.com/davidyyd/alpha-SVRG)

527、FaceShot: Bring Any Character into Life

- Paper: [**点击下载**](https://openreview.net/attachment?id=oJA1GUqRww&name=pdf)
- Code: [https://faceshot2024.github.io/faceshot/](https://faceshot2024.github.io/faceshot/)

528、Do Deep Neural Network Solutions Form a Star Domain?

- Paper: [**点击下载**](https://openreview.net/attachment?id=QjO0fUlVYK&name=pdf)
- Code: [https://github.com/aktsonthalia/starlight](https://github.com/aktsonthalia/starlight)

529、FakeShield: Explainable Image Forgery Detection and Localization via Multi-modal Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=pAQzEY7M03&name=pdf)
- Code: [https://github.com/zhipeixu/FakeShield](https://github.com/zhipeixu/FakeShield)

530、Revisiting In-context Learning Inference Circuit in Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=xizpnYNvQq&name=pdf)
- Code: [https://orcid.org/0000-0002-7127-1954](https://orcid.org/0000-0002-7127-1954)

531、VideoGrain: Modulating Space-Time Attention for Multi-Grained Video Editing

- Paper: [**点击下载**](https://openreview.net/attachment?id=SSslAtcPB6&name=pdf)
- Code: [https://knightyxp.github.io/VideoGrain_project_page](https://knightyxp.github.io/VideoGrain_project_page)

532、COMBO: Compositional World Models for Embodied Multi-Agent Cooperation

- Paper: [**点击下载**](https://openreview.net/attachment?id=YXRyYkb1im&name=pdf)
- Code: [https://embodied-agi.cs.umass.edu/combo/](https://embodied-agi.cs.umass.edu/combo/)

533、FasterCache: Training-Free Video Diffusion Model Acceleration with High Quality

- Paper: [**点击下载**](https://openreview.net/attachment?id=W49UjcpGxx&name=pdf)
- Code: [https://github.com/Vchitect/FasterCache](https://github.com/Vchitect/FasterCache)

534、Long-horizon Visual Instruction Generation with Logic and Attribute Self-reflection

- Paper: [**点击下载**](https://openreview.net/attachment?id=EdMb9TqqDY&name=pdf)
- Code: [https://github.com/suoych/LIGER](https://github.com/suoych/LIGER)

535、SaRA: High-Efficient Diffusion Model Fine-tuning with Progressive Sparse Low-Rank Adaptation

- Paper: [**点击下载**](https://openreview.net/attachment?id=wGVOxplEbf&name=pdf)
- Code: [https://sjtuplayer.github.io/projects/SaRA](https://sjtuplayer.github.io/projects/SaRA)

536、Mitigating Hallucination in Large Vision-Language Models via Modular Attribution and Intervention

- Paper: [**点击下载**](https://openreview.net/attachment?id=Bjq4W7P2Us&name=pdf)
- Code: [https://github.com/TianyunYoung/Hallucination-Attribution](https://github.com/TianyunYoung/Hallucination-Attribution)

537、Gated Delta Networks: Improving Mamba2 with Delta Rule

- Paper: [**点击下载**](https://openreview.net/attachment?id=r8H7xhYPwz&name=pdf)
- Code: [https://github.com/NVlabs/GatedDeltaNet](https://github.com/NVlabs/GatedDeltaNet)

538、3DTrajMaster: Mastering 3D Trajectory for Multi-Entity Motion in Video Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=Gx04TnVjee&name=pdf)
- Code: [http://fuxiao0719.github.io/projects/3dtrajmaster](http://fuxiao0719.github.io/projects/3dtrajmaster)

539、Understanding and Enhancing the Transferability of Jailbreaking Attacks

- Paper: [**点击下载**](https://openreview.net/attachment?id=asR9FVd4eL&name=pdf)
- Code: [https://github.com/tmllab/2025_ICLR_PiF](https://github.com/tmllab/2025_ICLR_PiF)

540、AniSDF: Fused-Granularity Neural Surfaces with Anisotropic Encoding for High-Fidelity 3D Reconstruction

- Paper: [**点击下载**](https://openreview.net/attachment?id=v1f6c7wVBm&name=pdf)
- Code: [https://g-1nonly.github.io/AniSDF_Website/](https://g-1nonly.github.io/AniSDF_Website/)

### Accept Oral

1、RM-Bench: Benchmarking Reward Models of Language Models with Subtlety and Style

- Paper: [**点击下载**](https://openreview.net/attachment?id=QEHrmQPBdd&name=pdf)
- Code: [https://github.com/THU-KEG/RM-Bench](https://github.com/THU-KEG/RM-Bench)

2、TopoLM: brain-like spatio-functional organization in a topographic language model

- Paper: [**点击下载**](https://openreview.net/attachment?id=aWXnKanInf&name=pdf)
- Code: [https://github.com/epflneuroailab/topolm](https://github.com/epflneuroailab/topolm)

3、Spider 2.0: Evaluating Language Models on Real-World Enterprise Text-to-SQL Workflows

- Paper: [**点击下载**](https://openreview.net/attachment?id=XmProj9cPs&name=pdf)
- Code: [spider2-sql.github.io](spider2-sql.github.io)

4、Diffusion-Based Planning for Autonomous Driving with Flexible Guidance

- Paper: [**点击下载**](https://openreview.net/attachment?id=wM2sfVgMDH&name=pdf)
- Code: [https://zhengyinan-air.github.io/Diffusion-Planner/](https://zhengyinan-air.github.io/Diffusion-Planner/)

5、Knowledge Entropy Decay during Language Model Pretraining Hinders New Knowledge Acquisition

- Paper: [**点击下载**](https://openreview.net/attachment?id=eHehzSDUFp&name=pdf)
- Code: [https://github.com/kaistAI/Knowledge-Entropy.git](https://github.com/kaistAI/Knowledge-Entropy.git)

6、Learning to Search from Demonstration Sequences

- Paper: [**点击下载**](https://openreview.net/attachment?id=v593OaNePQ&name=pdf)
- Code: [https://github.com/dixantmittal/differentiable-tree-search-network](https://github.com/dixantmittal/differentiable-tree-search-network)

7、Measuring and Enhancing Trustworthiness of LLMs in RAG through Grounded Attributions and Learning to Refuse

- Paper: [**点击下载**](https://openreview.net/attachment?id=Iyrtb9EJBp&name=pdf)
- Code: [https://github.com/declare-lab/trust-align](https://github.com/declare-lab/trust-align)

8、MAP: Multi-Human-Value Alignment Palette

- Paper: [**点击下载**](https://openreview.net/attachment?id=NN6QHwgRrQ&name=pdf)
- Code: [https://github.com/wang8740/MAP](https://github.com/wang8740/MAP)

9、Spread Preference Annotation: Direct Preference Judgment for Efficient LLM Alignment

- Paper: [**点击下载**](https://openreview.net/attachment?id=BPgK5XW1Nb&name=pdf)
- Code: [https://github.com/kingdy2002/SPA](https://github.com/kingdy2002/SPA)

10、Learning to Discover Regulatory Elements for Gene Expression Prediction

- Paper: [**点击下载**](https://openreview.net/attachment?id=Mfnh1Sqdwf&name=pdf)
- Code: [https://github.com/divelab/AIRS/](https://github.com/divelab/AIRS/)

11、Brain Bandit: A Biologically Grounded Neural Network for Efficient Control of Exploration

- Paper: [**点击下载**](https://openreview.net/attachment?id=RWJX5F5I9g&name=pdf)
- Code: [https://github.com/anonymousforICLR/BrainBandit](https://github.com/anonymousforICLR/BrainBandit)

12、Block Diffusion: Interpolating Between Autoregressive and Diffusion Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=tyEyYT267x&name=pdf)
- Code: [https://mariannearriola.github.io/bd3-lms/](https://mariannearriola.github.io/bd3-lms/)

13、Progressive Compression with Universally Quantized Diffusion Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=CxXGvKRDnL&name=pdf)
- Code: [https://github.com/mandt-lab/uqdm](https://github.com/mandt-lab/uqdm)

14、Standard Gaussian Process is All You Need for High-Dimensional Bayesian Optimization

- Paper: [**点击下载**](https://openreview.net/attachment?id=kX8h23UG6v&name=pdf)
- Code: [https://github.com/XZT008/Standard-GP-is-all-you-need-for-HDBO](https://github.com/XZT008/Standard-GP-is-all-you-need-for-HDBO)

15、Instant Policy: In-Context Imitation Learning via Graph Diffusion

- Paper: [**点击下载**](https://openreview.net/attachment?id=je3GZissZc&name=pdf)
- Code: [https://www.robot-learning.uk/instant-policy](https://www.robot-learning.uk/instant-policy)

16、RB-Modulation: Training-Free Stylization using Reference-Based Modulation

- Paper: [**点击下载**](https://openreview.net/attachment?id=bnINPG5A32&name=pdf)
- Code: [https://rb-modulation.github.io/](https://rb-modulation.github.io/)

17、Kinetix: Investigating the Training of General Agents through Open-Ended Physics-Based Control Tasks

- Paper: [**点击下载**](https://openreview.net/attachment?id=zCxGCdzreM&name=pdf)
- Code: [https://kinetix-env.github.io](https://kinetix-env.github.io)

18、Oscillatory State-Space Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=GRMfXcAAFh&name=pdf)
- Code: [https://github.com/tk-rusch/linoss](https://github.com/tk-rusch/linoss)

19、Attention as a Hypernetwork

- Paper: [**点击下载**](https://openreview.net/attachment?id=V4K9h1qNxE&name=pdf)
- Code: [https://github.com/smonsays/hypernetwork-attention](https://github.com/smonsays/hypernetwork-attention)

20、Energy-based Backdoor Defense Against Federated Graph Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=5Jc7r5aqHJ&name=pdf)
- Code: [https://github.com/ZitongShi/fedTGE](https://github.com/ZitongShi/fedTGE)

21、Toward Guidance-Free AR Visual Generation via Condition Contrastive Alignment

- Paper: [**点击下载**](https://openreview.net/attachment?id=kGvXIlIVLM&name=pdf)
- Code: [https://github.com/thu-ml/CCA](https://github.com/thu-ml/CCA)

22、RMP-SAM: Towards Real-Time Multi-Purpose Segment Anything

- Paper: [**点击下载**](https://openreview.net/attachment?id=1pXzC30ry5&name=pdf)
- Code: [https://github.com/xushilin1/RAP-SAM](https://github.com/xushilin1/RAP-SAM)

23、Residual Deep Gaussian Processes on Manifolds

- Paper: [**点击下载**](https://openreview.net/attachment?id=JWtrk7mprJ&name=pdf)
- Code: [https://github.com/KacperWyrwal/residual-deep-gps](https://github.com/KacperWyrwal/residual-deep-gps)

24、Learning to Discretize Denoising Diffusion ODEs

- Paper: [**点击下载**](https://openreview.net/attachment?id=xDrFWUmCne&name=pdf)
- Code: [https://github.com/vinhsuhi/LD3](https://github.com/vinhsuhi/LD3)

25、Proteina: Scaling Flow-based Protein Structure Generative Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=TVQLu34bdw&name=pdf)
- Code: [https://research.nvidia.com/labs/genair/proteina/](https://research.nvidia.com/labs/genair/proteina/)

26、Feedback Favors the Generalization of Neural ODEs

- Paper: [**点击下载**](https://openreview.net/attachment?id=cmfyMV45XO&name=pdf)
- Code: [https://sites.google.com/view/feedbacknn](https://sites.google.com/view/feedbacknn)

27、Scaling and evaluating sparse autoencoders

- Paper: [**点击下载**](https://openreview.net/attachment?id=tcsZt9ZNKD&name=pdf)
- Code: [http://example.com](http://example.com)

28、Do I Know This Entity? Knowledge Awareness and Hallucinations in Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=WCRQFlji2q&name=pdf)
- Code: [https://github.com/javiferran/sae_entities](https://github.com/javiferran/sae_entities)

29、TetSphere Splatting: Representing High-Quality Geometry with Lagrangian Volumetric Meshes

- Paper: [**点击下载**](https://openreview.net/attachment?id=8enWnd6Gp3&name=pdf)
- Code: [https://github.com/gmh14/tssplat](https://github.com/gmh14/tssplat)

30、Copyright-Protected Language Generation via Adaptive Model Fusion

- Paper: [**点击下载**](https://openreview.net/attachment?id=kRoWeLTpL4&name=pdf)
- Code: [https://github.com/jaabmar/cp_fuse](https://github.com/jaabmar/cp_fuse)

31、BIRD: A Trustworthy Bayesian Inference Framework for Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=fAAaT826Vv&name=pdf)
- Code: [https://github.com/CogComp/BIRD](https://github.com/CogComp/BIRD)

32、Rethinking Reward Modeling in Preference-based Large Language Model Alignment

- Paper: [**点击下载**](https://openreview.net/attachment?id=rfdblE10qm&name=pdf)
- Code: [https://github.com/holarissun/RewardModelingBeyondBradleyTerry](https://github.com/holarissun/RewardModelingBeyondBradleyTerry)

33、Progressive distillation induces an implicit curriculum

- Paper: [**点击下载**](https://openreview.net/attachment?id=wPMRwmytZe&name=pdf)
- Code: [https://github.com/abhishekpanigrahi1996/ProgressiveDistillation](https://github.com/abhishekpanigrahi1996/ProgressiveDistillation)

34、miniCTX: Neural Theorem Proving with (Long-)Contexts

- Paper: [**点击下载**](https://openreview.net/attachment?id=KIgaAqEFHW&name=pdf)
- Code: [https://cmu-l3.github.io/minictx](https://cmu-l3.github.io/minictx)

35、Ctrl-Adapter: An Efficient and Versatile Framework for Adapting Diverse Controls to Any Diffusion Model

- Paper: [**点击下载**](https://openreview.net/attachment?id=ny8T8OuNHe&name=pdf)
- Code: [https://ctrl-adapter.github.io](https://ctrl-adapter.github.io)

36、Predictive Inverse Dynamics Models are Scalable Learners for Robotic Manipulation

- Paper: [**点击下载**](https://openreview.net/attachment?id=meRCKuUpmc&name=pdf)
- Code: [https://nimolty.github.io/Seer/](https://nimolty.github.io/Seer/)

37、Representation Alignment for Generation: Training Diffusion Transformers Is Easier Than You Think

- Paper: [**点击下载**](https://openreview.net/attachment?id=DJSZGGZYVi&name=pdf)
- Code: [https://sihyun.me/REPA](https://sihyun.me/REPA)

38、MMQA: Evaluating LLMs with Multi-Table Multi-Hop Complex Questions

- Paper: [**点击下载**](https://openreview.net/attachment?id=GGlpykXDCa&name=pdf)
- Code: [https://anonymous.4open.science/r/MMQA-34B1](https://anonymous.4open.science/r/MMQA-34B1)

39、On Scaling Up 3D Gaussian Splatting Training

- Paper: [**点击下载**](https://openreview.net/attachment?id=pQqeQpMkE7&name=pdf)
- Code: [https://github.com/nyu-systems/Grendel-GS](https://github.com/nyu-systems/Grendel-GS)

40、BigCodeBench: Benchmarking Code Generation with Diverse Function Calls and Complex Instructions

- Paper: [**点击下载**](https://openreview.net/attachment?id=YrycTjllL0&name=pdf)
- Code: [https://bigcode-bench.github.io/](https://bigcode-bench.github.io/)

41、SD-LoRA: Scalable Decoupled Low-Rank Adaptation for Class Incremental Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=5U1rlpX68A&name=pdf)
- Code: [https://github.com/WuYichen-97/SD-Lora-CL](https://github.com/WuYichen-97/SD-Lora-CL)

42、Improving Probabilistic Diffusion Models With Optimal Diagonal Covariance Matching

- Paper: [**点击下载**](https://openreview.net/attachment?id=fV0t65OBUu&name=pdf)
- Code: [https://github.com/J-zin/OCM_DPM](https://github.com/J-zin/OCM_DPM)

43、PathGen-1.6M: 1.6 Million Pathology Image-text Pairs Generation through Multi-agent Collaboration

- Paper: [**点击下载**](https://openreview.net/attachment?id=rFpZnn11gj&name=pdf)
- Code: [https://github.com/PathFoundation/PathGen-1.6M](https://github.com/PathFoundation/PathGen-1.6M)

44、A Probabilistic Perspective on Unlearning and Alignment for Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=51WraMid8K&name=pdf)
- Code: [https://www.cs.cit.tum.de/daml/probabilistic-unlearning/](https://www.cs.cit.tum.de/daml/probabilistic-unlearning/)

45、MLE-bench: Evaluating Machine Learning Agents on Machine Learning Engineering

- Paper: [**点击下载**](https://openreview.net/attachment?id=6s5uXNWGIh&name=pdf)
- Code: [https://github.com/openai/mle-bench/](https://github.com/openai/mle-bench/)

46、Subgraph Federated Learning for Local Generalization

- Paper: [**点击下载**](https://openreview.net/attachment?id=cH65nS5sOz&name=pdf)
- Code: [https://github.com/sung-won-kim/FedLoG](https://github.com/sung-won-kim/FedLoG)

47、Amortized Control of Continuous State Space Feynman-Kac Model for Irregular Time Series

- Paper: [**点击下载**](https://openreview.net/attachment?id=8zJRon6k5v&name=pdf)
- Code: [https://github.com/bw-park/ACSSM](https://github.com/bw-park/ACSSM)

48、Cheating Automatic LLM Benchmarks: Null Models Achieve High Win Rates

- Paper: [**点击下载**](https://openreview.net/attachment?id=syThiTmWWm&name=pdf)
- Code: [https://github.com/sail-sg/Cheating-LLM-Benchmarks](https://github.com/sail-sg/Cheating-LLM-Benchmarks)

49、Combatting Dimensional Collapse in LLM Pre-Training Data via Submodular File Selection

- Paper: [**点击下载**](https://openreview.net/attachment?id=f4gF6AIHRy&name=pdf)
- Code: [https://github.com/MediaBrain-SJTU/DiSF.git](https://github.com/MediaBrain-SJTU/DiSF.git)

50、SymmetricDiffusers: Learning Discrete Diffusion Models over Finite Symmetric Groups

- Paper: [**点击下载**](https://openreview.net/attachment?id=EO8xpnW7aX&name=pdf)
- Code: [https://github.com/DSL-Lab/SymmetricDiffusers](https://github.com/DSL-Lab/SymmetricDiffusers)

51、Data Scaling Laws in Imitation Learning for Robotic Manipulation

- Paper: [**点击下载**](https://openreview.net/attachment?id=pISLZG7ktL&name=pdf)
- Code: [https://data-scaling-laws.github.io/](https://data-scaling-laws.github.io/)

52、Population Transformer: Learning Population-level Representations of Neural Activity

- Paper: [**点击下载**](https://openreview.net/attachment?id=FVuqJt3c4L&name=pdf)
- Code: [https://github.com/czlwang/PopulationTransformer](https://github.com/czlwang/PopulationTransformer)

53、HiRA: Parameter-Efficient Hadamard High-Rank Adaptation for Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=TwJrTz9cRS&name=pdf)
- Code: [https://github.com/hqsiswiliam/hira](https://github.com/hqsiswiliam/hira)

54、On Conformal Isometry of Grid Cells: Learning Distance-Preserving Position Embedding

- Paper: [**点击下载**](https://openreview.net/attachment?id=Xo0Q1N7CGk&name=pdf)
- Code: [https://github.com/DehongXu/grid-cell-conformal-isometry](https://github.com/DehongXu/grid-cell-conformal-isometry)

55、How much of my dataset did you use? Quantitative Data Usage Inference in Machine Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=EUSkm2sVJ6&name=pdf)
- Code: [https://github.com/privacytrustlab/ml_privacy_meter/tree/master/research](https://github.com/privacytrustlab/ml_privacy_meter/tree/master/research)

56、A Theoretically-Principled Sparse, Connected, and Rigid Graph Representation of Molecules

- Paper: [**点击下载**](https://openreview.net/attachment?id=OIvg3MqWX2&name=pdf)
- Code: [https://github.com/Utah-Math-Data-Science/UnitSphere](https://github.com/Utah-Math-Data-Science/UnitSphere)

57、MOS: Model Synergy for Test-Time Adaptation on LiDAR-Based 3D Object Detection

- Paper: [**点击下载**](https://openreview.net/attachment?id=Y6aHdDNQYD&name=pdf)
- Code: [https://github.com/zhuoxiao-chen/MOS](https://github.com/zhuoxiao-chen/MOS)

58、LARP: Tokenizing Videos with a Learned Autoregressive Generative Prior

- Paper: [**点击下载**](https://openreview.net/attachment?id=Wr3UuEx72f&name=pdf)
- Code: [https://hywang66.github.io/larp/](https://hywang66.github.io/larp/)

59、Unlocking the Power of Function Vectors for Characterizing and Mitigating Catastrophic Forgetting in Continual Instruction Tuning

- Paper: [**点击下载**](https://openreview.net/attachment?id=gc8QAQfXv6&name=pdf)
- Code: [https://github.com/GangweiJiang/FvForgetting.git](https://github.com/GangweiJiang/FvForgetting.git)

60、EmbodiedSAM: Online Segment Any 3D Thing in Real Time

- Paper: [**点击下载**](https://openreview.net/attachment?id=XFYUwIyTxQ&name=pdf)
- Code: [https://xuxw98.github.io/ESAM/](https://xuxw98.github.io/ESAM/)

61、Robustness Inspired Graph Backdoor Defense

- Paper: [**点击下载**](https://openreview.net/attachment?id=trKNi4IUiP&name=pdf)
- Code: [https://github.com/zzwjames/RIGBD](https://github.com/zzwjames/RIGBD)

62、Proxy Denoising for Source-Free Domain Adaptation

- Paper: [**点击下载**](https://openreview.net/attachment?id=FIj9IEPCKr&name=pdf)
- Code: [https://github.com/tntek/source-free-domain-adaptation](https://github.com/tntek/source-free-domain-adaptation)

63、Cybench: A Framework for Evaluating Cybersecurity Capabilities and Risks of Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=tc90LV0yRL&name=pdf)
- Code: [https://cybench.github.io](https://cybench.github.io)

64、On the Identification of Temporal Causal Representation with Instantaneous Dependence

- Paper: [**点击下载**](https://openreview.net/attachment?id=2efNHgYRvM&name=pdf)
- Code: [https://github.com/DMIRLAB-Group/IDOL](https://github.com/DMIRLAB-Group/IDOL)

65、WizardMath: Empowering Mathematical Reasoning for Large Language Models via Reinforced Evol-Instruct

- Paper: [**点击下载**](https://openreview.net/attachment?id=mMPMHWOdOy&name=pdf)
- Code: [https://github.com/nlpxucan/WizardLM](https://github.com/nlpxucan/WizardLM)

66、Learning Dynamics of LLM Finetuning

- Paper: [**点击下载**](https://openreview.net/attachment?id=tPNHOoZFl9&name=pdf)
- Code: [https://github.com/Joshua-Ren/Learning_dynamics_LLM](https://github.com/Joshua-Ren/Learning_dynamics_LLM)

67、More RLHF, More Trust? On The Impact of Preference Alignment On Trustworthiness

- Paper: [**点击下载**](https://openreview.net/attachment?id=FpiCLJrSW8&name=pdf)
- Code: [https://github.com/AI4LIFE-GROUP/RLHF_Trust](https://github.com/AI4LIFE-GROUP/RLHF_Trust)

68、Topological Blindspots: Understanding and Extending Topological Deep Learning Through the Lens of Expressivity

- Paper: [**点击下载**](https://openreview.net/attachment?id=EzjsoomYEb&name=pdf)
- Code: [https://github.com/yoavgelberg/SMCN](https://github.com/yoavgelberg/SMCN)

69、Geometry-aware RL for Manipulation of Varying Shapes and Deformable Objects

- Paper: [**点击下载**](https://openreview.net/attachment?id=7BLXhmWvwF&name=pdf)
- Code: [https://thobotics.github.io/hepi](https://thobotics.github.io/hepi)

70、ProtComposer: Compositional Protein Structure Generation with 3D Ellipsoids

- Paper: [**点击下载**](https://openreview.net/attachment?id=0ctvBgKFgc&name=pdf)
- Code: [https://github.com/NVlabs/protcomposer](https://github.com/NVlabs/protcomposer)

71、MoE++: Accelerating Mixture-of-Experts Methods with Zero-Computation Experts

- Paper: [**点击下载**](https://openreview.net/attachment?id=t7P5BUKcYv&name=pdf)
- Code: [https://github.com/SkyworkAI/MoE-plus-plus](https://github.com/SkyworkAI/MoE-plus-plus)

72、Compositional Entailment Learning for Hyperbolic Vision-Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=3i13Gev2hV&name=pdf)
- Code: [https://github.com/PalAvik/hycoclip](https://github.com/PalAvik/hycoclip)

73、DeepLTL: Learning to Efficiently Satisfy Complex LTL Specifications for Multi-Task RL

- Paper: [**点击下载**](https://openreview.net/attachment?id=9pW2J49flQ&name=pdf)
- Code: [https://deep-ltl.github.io/](https://deep-ltl.github.io/)

74、Composing Unbalanced Flows for Flexible Docking and Relaxation

- Paper: [**点击下载**](https://openreview.net/attachment?id=gHLWTzKiZV&name=pdf)
- Code: [https://github.com/vsomnath/flexdock](https://github.com/vsomnath/flexdock)

75、On the Role of Attention Heads in Large Language Model Safety

- Paper: [**点击下载**](https://openreview.net/attachment?id=h0Ak8A5yqw&name=pdf)
- Code: [https://github.com/ydyjya/SafetyHeadAttribution](https://github.com/ydyjya/SafetyHeadAttribution)

76、AlphaEdit: Null-Space Constrained Model Editing for Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=HvSytvg3Jh&name=pdf)
- Code: [https://github.com/jianghoucheng/AlphaEdit](https://github.com/jianghoucheng/AlphaEdit)

77、Prioritized Generative Replay

- Paper: [**点击下载**](https://openreview.net/attachment?id=5IkDAfabuo&name=pdf)
- Code: [https://github.com/renwang435/pgr](https://github.com/renwang435/pgr)

78、Learning stochastic dynamics from snapshots through regularized unbalanced optimal transport

- Paper: [**点击下载**](https://openreview.net/attachment?id=gQlxd3Mtru&name=pdf)
- Code: [https://github.com/zhenyiizhang/DeepRUOT](https://github.com/zhenyiizhang/DeepRUOT)

79、No Pose, No Problem: Surprisingly Simple 3D Gaussian Splats from Sparse Unposed Images

- Paper: [**点击下载**](https://openreview.net/attachment?id=P4o9akekdf&name=pdf)
- Code: [https://noposplat.github.io](https://noposplat.github.io)

80、The Geometry of Categorical and Hierarchical Concepts in Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=bVTM2QKYuA&name=pdf)
- Code: [https://github.com/KihoPark/LLM_Categorical_Hierarchical_Representations](https://github.com/KihoPark/LLM_Categorical_Hierarchical_Representations)

81、Variational Diffusion Posterior Sampling with Midpoint Guidance

- Paper: [**点击下载**](https://openreview.net/attachment?id=6EUtjXAvmj&name=pdf)
- Code: [https://github.com/yazidjanati/mgps](https://github.com/yazidjanati/mgps)

82、Retrieval Head Mechanistically Explains Long-Context Factuality

- Paper: [**点击下载**](https://openreview.net/attachment?id=EytBpUGB1Z&name=pdf)
- Code: [https://github.com/nightdessert/Retrieval_Head](https://github.com/nightdessert/Retrieval_Head)

83、NeuralPlane: Structured 3D Reconstruction in Planar Primitives with Neural Fields

- Paper: [**点击下载**](https://openreview.net/attachment?id=5UKrnKuspb&name=pdf)
- Code: [https://neuralplane.github.io/](https://neuralplane.github.io/)

84、Differential Transformer

- Paper: [**点击下载**](https://openreview.net/attachment?id=OvoCm1gGhN&name=pdf)
- Code: [https://aka.ms/GeneralAI](https://aka.ms/GeneralAI)

85、High-Dynamic Radar Sequence Prediction for Weather Nowcasting Using Spatiotemporal Coherent Gaussian Representation

- Paper: [**点击下载**](https://openreview.net/attachment?id=Cjz9Xhm7sI&name=pdf)
- Code: [https://ziyeeee.github.io/stcgs.github.io/](https://ziyeeee.github.io/stcgs.github.io/)

86、Trust or Escalate: LLM Judges with Provable Guarantees for Human Agreement

- Paper: [**点击下载**](https://openreview.net/attachment?id=UHPnqSTBPO&name=pdf)
- Code: [https://github.com/jaehunjung1/cascaded-selective-evaluation](https://github.com/jaehunjung1/cascaded-selective-evaluation)

87、REEF: Representation Encoding Fingerprints for Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=SnDmPkOJ0T&name=pdf)
- Code: [https://github.com/AI45Lab/REEF](https://github.com/AI45Lab/REEF)

88、Your Mixture-of-Experts LLM Is Secretly an Embedding Model for Free

- Paper: [**点击下载**](https://openreview.net/attachment?id=eFGQ97z5Cd&name=pdf)
- Code: [https://github.com/tianyi-lab/MoE-Embedding](https://github.com/tianyi-lab/MoE-Embedding)

89、Data Selection via Optimal Control for Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=dhAL5fy8wS&name=pdf)
- Code: [https://github.com/microsoft/LMOps/tree/main/data_selection](https://github.com/microsoft/LMOps/tree/main/data_selection)

90、GridMix: Exploring Spatial Modulation for Neural Fields in PDE Modeling

- Paper: [**点击下载**](https://openreview.net/attachment?id=Fur0DtynPX&name=pdf)
- Code: [https://github.com/LeapLabTHU/GridMix](https://github.com/LeapLabTHU/GridMix)

91、LLM-SR: Scientific Equation Discovery via Programming with Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=m2nmp8P5in&name=pdf)
- Code: [https://github.com/deep-symbolic-mathematics/LLM-SR](https://github.com/deep-symbolic-mathematics/LLM-SR)

92、Rethinking the generalization of drug target affinity prediction algorithms via similarity aware evaluation

- Paper: [**点击下载**](https://openreview.net/attachment?id=j7cyANIAxV&name=pdf)
- Code: [https://github.com/Amshoreline/SAE/tree/main](https://github.com/Amshoreline/SAE/tree/main)

93、LVSM: A Large View Synthesis Model with Minimal 3D Inductive Bias

- Paper: [**点击下载**](https://openreview.net/attachment?id=QQBPWtvtcn&name=pdf)
- Code: [https://haian-jin.github.io/projects/LVSM/](https://haian-jin.github.io/projects/LVSM/)

94、Can a MISL Fly? Analysis and Ingredients for Mutual Information Skill Learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=xoIeVdFO7U&name=pdf)
- Code: [https://princeton-rl.github.io/contrastive-successor-features](https://princeton-rl.github.io/contrastive-successor-features)

95、Cut Your Losses in Large-Vocabulary Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=E4Fk3YuG56&name=pdf)
- Code: [https://github.com/apple/ml-cross-entropy](https://github.com/apple/ml-cross-entropy)

96、Two Effects, One Trigger: On the Modality Gap, Object Bias, and Information Imbalance in Contrastive Vision-Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=uAFHCZRmXk&name=pdf)
- Code: [https://github.com/lmb-freiburg/two-effects-one-trigger](https://github.com/lmb-freiburg/two-effects-one-trigger)

97、FlexPrefill: A Context-Aware Sparse Attention Mechanism for Efficient Long-Sequence Inference

- Paper: [**点击下载**](https://openreview.net/attachment?id=OfjIlbelrT&name=pdf)
- Code: [https://github.com/bytedance/FlexPrefill](https://github.com/bytedance/FlexPrefill)

98、SANA: Efficient High-Resolution Text-to-Image Synthesis with Linear Diffusion Transformers

- Paper: [**点击下载**](https://openreview.net/attachment?id=N8Oj1XhtYZ&name=pdf)
- Code: [https://nvlabs.github.io/Sana](https://nvlabs.github.io/Sana)

99、REGENT: A Retrieval-Augmented Generalist Agent That Can Act In-Context in New Environments

- Paper: [**点击下载**](https://openreview.net/attachment?id=NxyfSW6mLK&name=pdf)
- Code: [https://kaustubhsridhar.github.io/regent-research](https://kaustubhsridhar.github.io/regent-research)

100、MMIE: Massive Multimodal Interleaved Comprehension Benchmark for Large Vision-Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=HnhNRrLPwm&name=pdf)
- Code: [https://mmie-bench.github.io/](https://mmie-bench.github.io/)

101、Do as We Do, Not as You Think: the Conformity of Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=st77ShxP1K&name=pdf)
- Code: [https://github.com/Zhiyuan-Weng/BenchForm](https://github.com/Zhiyuan-Weng/BenchForm)

102、Booster: Tackling Harmful Fine-tuning for Large Language Models via Attenuating Harmful Perturbation

- Paper: [**点击下载**](https://openreview.net/attachment?id=tTPHgb0EtV&name=pdf)
- Code: [https://github.com/git-disl/Booster](https://github.com/git-disl/Booster)

103、Probabilistic Learning to Defer: Handling Missing Expert Annotations and Controlling Workload Distribution

- Paper: [**点击下载**](https://openreview.net/attachment?id=zl0HLZOJC9&name=pdf)
- Code: [https://orcid.org/0000-0003-2672-6291](https://orcid.org/0000-0003-2672-6291)

104、Artificial Kuramoto Oscillatory Neurons

- Paper: [**点击下载**](https://openreview.net/attachment?id=nwDRD4AMoN&name=pdf)
- Code: [https://github.com/autonomousvision/akorn](https://github.com/autonomousvision/akorn)

105、ChartMoE: Mixture of Diversely Aligned Expert Connector for Chart Understanding

- Paper: [**点击下载**](https://openreview.net/attachment?id=o5TsWTUSeF&name=pdf)
- Code: [https://github.com/IDEA-FinAI/ChartMoE](https://github.com/IDEA-FinAI/ChartMoE)

106、A Decade's Battle on Dataset Bias: Are We There Yet?

- Paper: [**点击下载**](https://openreview.net/attachment?id=SctfBCLmWo&name=pdf)
- Code: [http://github.com/liuzhuang13/bias](http://github.com/liuzhuang13/bias)

107、Knowing Your Target: Target-Aware Transformer Makes Better Spatio-Temporal Video Grounding

- Paper: [**点击下载**](https://openreview.net/attachment?id=WOzffPgVjF&name=pdf)
- Code: [https://github.com/HengLan/TA-STVG](https://github.com/HengLan/TA-STVG)

108、Open-World Reinforcement Learning over Long Short-Term Imagination

- Paper: [**点击下载**](https://openreview.net/attachment?id=vzItLaEoDa&name=pdf)
- Code: [https://qiwang067.github.io/ls-imagine](https://qiwang067.github.io/ls-imagine)

109、SAM 2: Segment Anything in Images and Videos

- Paper: [**点击下载**](https://openreview.net/attachment?id=Ha6RTeWMd0&name=pdf)
- Code: [https://github.com/facebookresearch/sam2](https://github.com/facebookresearch/sam2)

110、Do Vision-Language Models Represent Space and How? Evaluating Spatial Frame of Reference under Ambiguities

- Paper: [**点击下载**](https://openreview.net/attachment?id=84pDoCD4lH&name=pdf)
- Code: [https://spatial-comfort.github.io/](https://spatial-comfort.github.io/)

111、A Computational Framework for Modeling Emergence of Color Vision in the Human Brain

- Paper: [**点击下载**](https://openreview.net/attachment?id=g3xuCtrG6H&name=pdf)
- Code: [https://matisse.eecs.berkeley.edu](https://matisse.eecs.berkeley.edu)

### Accept conditionaloral

1、Navigating the Digital World as Humans Do: Universal Visual Grounding for GUI Agents

- Paper: [**点击下载**](https://openreview.net/attachment?id=kxnoqaisCT&name=pdf)
- Code: [https://osu-nlp-group.github.io/UGround/](https://osu-nlp-group.github.io/UGround/)

2、TANGO: Co-Speech Gesture Video Reenactment with Hierarchical Audio Motion Embedding and Diffusion Interpolation

- Paper: [**点击下载**](https://openreview.net/attachment?id=LbEWwJOufy&name=pdf)
- Code: [https://github.com/CyberAgentAILab/TANGO](https://github.com/CyberAgentAILab/TANGO)

3、Loopy: Taming Audio-Driven Portrait Avatar with Long-Term Motion Dependency

- Paper: [**点击下载**](https://openreview.net/attachment?id=weM4YBicIP&name=pdf)
- Code: [https://loopyavataranony.github.io/](https://loopyavataranony.github.io/)

4、CyberHost: A One-stage Diffusion Framework for Audio-driven Talking Body Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=vaEPihQsAA&name=pdf)
- Code: [https://cyberhost.github.io/](https://cyberhost.github.io/)

5、PhysBench: Benchmarking and Enhancing Vision-Language Models for Physical World Understanding

- Paper: [**点击下载**](https://openreview.net/attachment?id=Q6a9W6kzv5&name=pdf)
- Code: [https://physbench.github.io/](https://physbench.github.io/)

### Accept conditionalspotlight

1、OmniCorpus: A Unified Multimodal Corpus of 10 Billion-Level Images Interleaved with Text

- Paper: [**点击下载**](https://openreview.net/attachment?id=kwqhn2VuG4&name=pdf)
- Code: [https://github.com/OpenGVLab/OmniCorpus](https://github.com/OpenGVLab/OmniCorpus)

2、FairMT-Bench: Benchmarking Fairness for Multi-turn Dialogue in Conversational LLMs

- Paper: [**点击下载**](https://openreview.net/attachment?id=RSGoXnS9GH&name=pdf)
- Code: [https://github.com/FanZT6/FairMT-bench](https://github.com/FanZT6/FairMT-bench)

### Accept conditionalposter

1、Scaling Speech-Text Pre-training with Synthetic Interleaved Data

- Paper: [**点击下载**](https://openreview.net/attachment?id=3tukjsVyrE&name=pdf)
- Code: [https://github.com/THUDM/GLM-4-Voice](https://github.com/THUDM/GLM-4-Voice)

2、K-HALU: Multiple Answer Korean Hallucination Benchmark for Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=VnLhUogHYE&name=pdf)
- Code: [https://github.com/J-Seo/K-HALU](https://github.com/J-Seo/K-HALU)

3、DiscoveryBench: Towards Data-Driven Discovery with Large Language Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=vyflgpwfJW&name=pdf)
- Code: [https://github.com/allenai/discoverybench](https://github.com/allenai/discoverybench)

4、Image Watermarks are Removable using Controllable Regeneration from Clean Noise

- Paper: [**点击下载**](https://openreview.net/attachment?id=mDKxlfraAn&name=pdf)
- Code: [https://github.com/yepengliu/CtrlRegen](https://github.com/yepengliu/CtrlRegen)

5、ScienceAgentBench: Toward Rigorous Assessment of Language Agents for Data-Driven Scientific Discovery

- Paper: [**点击下载**](https://openreview.net/attachment?id=6z4YKr0GK6&name=pdf)
- Code: [https://osu-nlp-group.github.io/ScienceAgentBench/](https://osu-nlp-group.github.io/ScienceAgentBench/)

6、Resolution Attack: Exploiting Image Compression to Deceive Deep Neural Networks

- Paper: [**点击下载**](https://openreview.net/attachment?id=OFukl9Qg8P&name=pdf)
- Code: [https://github.com/ywj1/resolution-attack](https://github.com/ywj1/resolution-attack)

7、SafeWatch: An Efficient Safety-Policy Following Video Guardrail Model with Transparent Explanations

- Paper: [**点击下载**](https://openreview.net/attachment?id=xjKz6IxgCX&name=pdf)
- Code: [https://safewatch-aiguard.github.io](https://safewatch-aiguard.github.io)

8、Interpretable Compressed Descriptions For Image Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=aXwukBD6M6&name=pdf)
- Code: [https://github.com/ArmandCom/InCoDe](https://github.com/ArmandCom/InCoDe)

9、Measuring And Improving Engagement of Text-to-Image Generation Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=TmCcNuo03f&name=pdf)
- Code: [behavior-in-the-wild.github.io/image-engagement](behavior-in-the-wild.github.io/image-engagement)

10、AgentHarm: A Benchmark for Measuring Harmfulness of LLM Agents

- Paper: [**点击下载**](https://openreview.net/attachment?id=AC5n7xHuR1&name=pdf)
- Code: [https://huggingface.co/datasets/ai-safety-institute/AgentHarm](https://huggingface.co/datasets/ai-safety-institute/AgentHarm)

11、Empowering LLM Agents with Zero-Shot Optimal Decision-Making through Q-learning

- Paper: [**点击下载**](https://openreview.net/attachment?id=JsVIGVntnQ&name=pdf)
- Code: [http://mlaq.site/](http://mlaq.site/)

12、TOMATO: Assessing Visual Temporal Reasoning Capabilities in Multimodal Foundation Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=fCi4o83Mfs&name=pdf)
- Code: [https://github.com/yale-nlp/TOMATO](https://github.com/yale-nlp/TOMATO)

13、Diversity-Rewarded CFG Distillation

- Paper: [**点击下载**](https://openreview.net/attachment?id=lJ66m0ibQL&name=pdf)
- Code: [https://google-research.github.io/seanet/musiclm/diverse_music/](https://google-research.github.io/seanet/musiclm/diverse_music/)

14、GUI-World: A Video Benchmark and Dataset for Multimodal GUI-oriented Understanding

- Paper: [**点击下载**](https://openreview.net/attachment?id=QarKTT5brZ&name=pdf)
- Code: [https://gui-world.github.io](https://gui-world.github.io)

15、Looking Backward: Streaming Video-to-Video Translation with Feature Banks

- Paper: [**点击下载**](https://openreview.net/attachment?id=AMkf7h7HER&name=pdf)
- Code: [https://jeff-liangf.github.io/projects/streamv2v](https://jeff-liangf.github.io/projects/streamv2v)

16、Personalized Representation from Personalized Generation

- Paper: [**点击下载**](https://openreview.net/attachment?id=jw7P4MHLWw&name=pdf)
- Code: [https://personalized-rep.github.io/](https://personalized-rep.github.io/)

17、VD3D: Taming Large Video Diffusion Transformers for 3D Camera Control

- Paper: [**点击下载**](https://openreview.net/attachment?id=0n4bS0R5MM&name=pdf)
- Code: [https://snap-research.github.io/vd3d](https://snap-research.github.io/vd3d)

18、Less is More: Masking Elements in Image Condition Features Avoids Content Leakages in Style Transfer Diffusion Models

- Paper: [**点击下载**](https://openreview.net/attachment?id=88JJjsLtqr&name=pdf)
- Code: [https://github.com/LinLLLL/MaskST](https://github.com/LinLLLL/MaskST)

19、$F^3Set$: Towards Analyzing Fast, Frequent, and Fine-grained Events from Videos

- Paper: [**点击下载**](https://openreview.net/attachment?id=vlg5WRKHxh&name=pdf)
- Code: [https://github.com/F3Set/F3Set](https://github.com/F3Set/F3Set)

20、SWEb: A Large Web Dataset for the Scandinavian Languages

- Paper: [**点击下载**](https://openreview.net/attachment?id=vhPE3PtTgC&name=pdf)
- Code: [https://huggingface.co/datasets/AI-Sweden-Models/SWEb](https://huggingface.co/datasets/AI-Sweden-Models/SWEb)

21、Unleashing the Potential of Vision-Language Pre-Training for 3D Zero-Shot Lesion Segmentation via Mask-Attribute Alignment

- Paper: [**点击下载**](https://openreview.net/attachment?id=QG31By6S6w&name=pdf)
- Code: [https://github.com/Yankai96/Malenia](https://github.com/Yankai96/Malenia)

22、PersonalLLM: Tailoring LLMs to Individual Preferences

- Paper: [**点击下载**](https://openreview.net/attachment?id=2R7498e2Tx&name=pdf)
- Code: [https://huggingface.co/datasets/namkoong-lab/PersonalLLM](https://huggingface.co/datasets/namkoong-lab/PersonalLLM)

23、BadRobot: Jailbreaking Embodied LLMs in the Physical World

- Paper: [**点击下载**](https://openreview.net/attachment?id=ei3qCntB66&name=pdf)
- Code: [https://Embodied-LLMs-Safety.github.io](https://Embodied-LLMs-Safety.github.io)

24、Grounding Multimodal Large Language Model in GUI World

- Paper: [**点击下载**](https://openreview.net/attachment?id=M9iky9Ruhx&name=pdf)
- Code: [https://github.com/showlab/AssistGUIGround](https://github.com/showlab/AssistGUIGround)

25、Personalized Visual Instruction Tuning

- Paper: [**点击下载**](https://openreview.net/attachment?id=sAxdIJ4l6z&name=pdf)
- Code: [https://github.com/sterzhang/PVIT](https://github.com/sterzhang/PVIT)

26、TaskGalaxy: Scaling Multi-modal Instruction Fine-tuning with Tens of Thousands Vision Task Types

- Paper: [**点击下载**](https://openreview.net/attachment?id=JXgnnUC0PH&name=pdf)
- Code: [https://github.com/Kwai-YuanQi/TaskGalaxy](https://github.com/Kwai-YuanQi/TaskGalaxy)

27、TAU-106K: A New Dataset for Comprehensive Understanding of Traffic Accident

- Paper: [**点击下载**](https://openreview.net/attachment?id=Fb0q2uI4Ha&name=pdf)
- Code: [https://github.com/cool-xuan/TABot](https://github.com/cool-xuan/TABot)

28、Learning to Generate Diverse Pedestrian Movements from Web Videos with Noisy Labels

- Paper: [**点击下载**](https://openreview.net/attachment?id=DydCqKa6AH&name=pdf)
- Code: [https://genforce.github.io/PedGen/](https://genforce.github.io/PedGen/)

29、UNIP: Rethinking Pre-trained Attention Patterns for Infrared Semantic Segmentation

- Paper: [**点击下载**](https://openreview.net/attachment?id=Xq7gwsnhPT&name=pdf)
- Code: [https://github.com/casiatao/UNIP](https://github.com/casiatao/UNIP)

30、DiTTo-TTS: Diffusion Transformers for Scalable Text-to-Speech without Domain-Specific Factors

- Paper: [**点击下载**](https://openreview.net/attachment?id=hQvX9MBowC&name=pdf)
- Code: [https://ditto-tts.github.io](https://ditto-tts.github.io)

31、Aria-MIDI: A Dataset of Piano MIDI Files for Symbolic Music Modeling

- Paper: [**点击下载**](https://openreview.net/attachment?id=X5hrhgndxW&name=pdf)
- Code: [https://github.com/loubbrad/aria-midi](https://github.com/loubbrad/aria-midi)

32、Depth Any Video with Scalable Synthetic Data

- Paper: [**点击下载**](https://openreview.net/attachment?id=gWqFbnKsqR&name=pdf)
- Code: [https://depthanyvideo.github.io](https://depthanyvideo.github.io)

33、HART: Efficient Visual Generation with Hybrid Autoregressive Transformer

- Paper: [**点击下载**](https://openreview.net/attachment?id=q5sOv4xQe4&name=pdf)
- Code: [https://hanlab.mit.edu/projects/hart](https://hanlab.mit.edu/projects/hart)

34、SONICS: Synthetic Or Not - Identifying Counterfeit Songs

- Paper: [**点击下载**](https://openreview.net/attachment?id=PY7KSh29Z8&name=pdf)
- Code: [https://github.com/awsaf49/sonics](https://github.com/awsaf49/sonics)

35、CycleResearcher: Improving Automated Research via Automated Review

- Paper: [**点击下载**](https://openreview.net/attachment?id=bjcsVLoHYs&name=pdf)
- Code: [https://wengsyx.github.io/Researcher/](https://wengsyx.github.io/Researcher/)

36、Interpretable Bilingual Multimodal Large Language Model for Diverse Biomedical Tasks

- Paper: [**点击下载**](https://openreview.net/attachment?id=YuHQTo6G9S&name=pdf)
- Code: [https://medrega.github.io](https://medrega.github.io)

ICLR2025-Papers-with-Code

ICLR2025-Papers-with-Code Latest Version Download

Today's picks

Was this content helpful?

Recent Feedback

ICLR2025-Papers-with-Code

ICLR2025-Papers-with-Code Latest Version Download

Today's picks

Was this content helpful?

Recent Feedback

Download README