user-as-engram
A user's facts become a few rows in a content-addressed memory table — not a rewrite of the model. Paper, code, data, and interactive site for "User as Engram" (arXiv:2606.19172).
파일 탐색기
- ANONYMOUS_NOTICE
- README.md
- __init__.py
- agent.py
- diagnose_v2.py
- diagnose_v3.py
- filter.py
- pipeline.sh
- probe_recite.py
- rewriter.py
- run_agent.py
- run_stage_a_v2.sh
- run_stage_a_v3.sh
- to_stage_a_format.py
- vector_db.py
- __init__.py
- aggregate_results.py
- baseline_icl.py
- leakage_test.py
- README.md
- stage_a.py
- stage_a_recite.py
- stage_b.py
- stage_c.py
- stage_c_pilot.py
- synth_users.py
- synth_users_medical.py
- __init__.py
- checkpoint_manager.py
- common.py
- core_eval.py
- dataloader.py
- dataset.py
- engine.py
- engram_module.py
- execution.py
- flash_attention.py
- fp8.py
- gpt.py
- logo.svg
- loss_eval.py
- optim.py
- report.py
- tokenizer.py
- ui.html
- base_eval.py
- base_train.py
- chat_cli.py
- chat_eval.py
- chat_rl.py
- chat_sft.py
- chat_web.py
- tok_eval.py
- tok_train.py
- arc.py
- common.py
- customjson.py
- gsm8k.py
- humaneval.py
- mmlu.py
- smoltalk.py
- spellingbee.py
- test_attention_fallback.py
- test_engine.py
- .gitignore
- .python-version
- LICENSE
- pyproject.toml
- README.md
- UPSTREAM.md
- uv.lock
- __init__.py
- additive_composition.py
- build_corpus.py
- build_corpus_xl.py
- build_corpus_xxl.py
- capacity_ablation_table.py
- cross_lm_transfer.py
- density_layered.py
- engram_finetune_mf.py
- engram_pretrain.py
- engram_server.py
- eval_at_scale.py
- eval_serving.py
- evaluate_replication.py
- factscale_table.py
- generate_opd_corpus.py
- generate_trace_corpus.py
- head_to_head_locality.py
- insertion_strategies_v2.py
- joint_opt.py
- judge_layered.py
- judge_locomo.py
- judge_only.py
- layered_architecture.py
- layered_rag.py
- layered_rag_scale.py
- locomo_eval.py
- longform_gen.py
- mech_depth.py
- mech_glassbox.py
- mech_lora_vs_engram.py
- mechanistic_analysis.py
- memory_systems_comparison.py
- memory_systems_paraphrase.py
- memory_systems_proper.py
- multifact_lora.py
- multihop_probe.py
- multihop_rag.py
- opt_strong_density.py
- paraphrase_test.py
- per_user_table_eval.py
- pick_optimal.py
- plot_mechanistic.py
- plot_runs.py
- qwen_rag_indirect.py
- qwen_rag_scale.py
- scalability_benchmark.py
- scaling_summary.py
- sft_baseline.py
- sft_engram_minimal.py
- train_shared_lora.py
- train_shared_lora_trace.py
- user_facts_demo.py
- README.md
- requirements.txt
- u000.json
- u001.json
- u002.json
- u003.json
- u004.json
- u005.json
- u006.json
- u007.json
- u008.json
- u009.json
- u010.json
- u011.json
- u012.json
- u013.json
- u014.json
- u015.json
- u016.json
- u017.json
- u018.json
- u019.json
- u020.json
- u021.json
- u022.json
- u023.json
- u024.json
- u025.json
- u026.json
- u027.json
- u028.json
- u029.json
- m000.json
- m001.json
- m002.json
- m003.json
- m004.json
- m005.json
- m006.json
- m007.json
- m008.json
- m009.json
- m010.json
- m011.json
- m012.json
- m013.json
- m014.json
- m015.json
- m016.json
- m017.json
- m018.json
- m019.json
- m020.json
- m021.json
- m022.json
- m023.json
- m024.json
- m025.json
- m026.json
- m027.json
- m028.json
- m029.json
- corpora.json
- corpora_xl.json
- corpora_xxl.json
- multihop_chains.json
- README.md
- chain_layered.sh
- chain_multischema.sh
- chain_opd.sh
- chain_rank_scaling.sh
- chain_reverse_xschema.sh
- gpu_supervisor.sh
- launch_engram_run.sh
- queue_experiments.sh
- run_capacity_ablation.sh
- run_chinchilla_chain.sh
- run_eval_suite.sh
- run_factscale_bench.sh
- run_full_analysis.sh
- run_larger_engrams.sh
- run_layered.sh
- run_multischema.sh
- run_opd.sh
- run_phase2_d12_1280.sh
- run_phase4_locomo_full.sh
- run_phaseH.sh
- run_rank_scaling.sh
- run_resume.sh
- run_reverse_xschema.sh
- run_scaling_and_h2h.sh
- run_supervisor.sh
- run_tier1_1a.sh
- run_tier2_6.sh
- probe_collisions.py
- probe_hash.py
- probe_readwrite.py
- fig1_layered_arch.py
- fig1_paradigm.py
- fig_density.py
- fig_pipeline.py
- fig_robustness.py
- relocate.py
- COMPREHENSIVE_FINDINGS.md
- FINDINGS.md
- LAYERED_ARCHITECTURE_PLAN.md
- MECH_FINDINGS_NEW.md
- MERGED_PAPER_OUTLINE.md
- notes.md
- PAPER.md
- PAPER_REFRAME_DRAFT.md
- README.md
- RESULTS.md
- roadmap.md
- TIER1_TIER2_PLAN.md
- tier2_plan.md
- fig1_paradigm.pdf
- fig1a_lora_recall.pdf
- fig1b_lora_peruser.pdf
- fig2_locality.pdf
- fig3_density_curves.pdf
- fig4_storage_scaling.pdf
- fig5_serving_latency.pdf
- fig6_logitlens.pdf
- fig_arch.pdf
- fig_capacity_heatmap.pdf
- fig_contamination.pdf
- fig_cross_base.pdf
- fig_crossbase.pdf
- fig_crossschema.pdf
- fig_dense_scaling.pdf
- fig_density.pdf
- fig_engram_in_transformer.pdf
- fig_engram_in_transformer.png
- fig_factscale.pdf
- fig_glassbox.pdf
- fig_joint_opt_dense.pdf
- fig_joint_opt_loss.pdf
- fig_landscape.pdf
- fig_layered_arch.pdf
- fig_layered_conditions.pdf
- fig_locomo_categories.pdf
- fig_locomo_judge.pdf
- fig_locomo_scaling.pdf
- fig_lora_vs_engram.pdf
- fig_memsystems.pdf
- fig_memsystems_a.pdf
- fig_memsystems_b.pdf
- fig_metric_mismatch.pdf
- fig_mf_finetune.pdf
- fig_multidomain.pdf
- fig_multihop.pdf
- fig_paraphrase.pdf
- fig_pareto.pdf
- fig_pareto_layered.pdf
- fig_pareto_rag.pdf
- fig_pareto_rag.png
- fig_pipeline.pdf
- fig_pretrain_loss.pdf
- fig_rag_scale.pdf
- fig_rag_scale.png
- fig_robustness.pdf
- fig_serving_cdf.pdf
- fig_serving_scale.pdf
- fig_shared_rank.pdf
- fig_strategy_bars.pdf
- fig_user_insertion.pdf
- fig_user_insertion.png
- abstract.txt
- arxiv.sty
- arxiv_submission.tar.gz
- fig_glassbox.py
- fig_landscape.py
- fig_layered_arch.py
- fig_locomo_categories.py
- fig_method_arch.py
- fig_pareto_rag.py
- fig_rag_scale.py
- gen_reorg_figs.py
- LICENSE
- main.pdf
- main.tex
- make_all_figures.sh
- make_figures.py
- make_figures_appendix.py
- make_new_figures.py
- README.md
- requirements.txt
- fig1_paradigm.pdf
- fig1a_lora_recall.pdf
- fig1b_lora_peruser.pdf
- fig2_locality.pdf
- fig3_density_curves.pdf
- fig4_storage_scaling.pdf
- fig5_serving_latency.pdf
- fig6_logitlens.pdf
- fig_arch.pdf
- fig_capacity_heatmap.pdf
- fig_contamination.pdf
- fig_cross_base.pdf
- fig_crossbase.pdf
- fig_crossschema.pdf
- fig_dense_scaling.pdf
- fig_density.pdf
- fig_engram_in_transformer.pdf
- fig_engram_in_transformer.png
- fig_factscale.pdf
- fig_glassbox.pdf
- fig_joint_opt_dense.pdf
- fig_joint_opt_loss.pdf
- fig_landscape.pdf
- fig_layered_arch.pdf
- fig_layered_arch.png
- fig_layered_conditions.pdf
- fig_locomo_categories.pdf
- fig_locomo_judge.pdf
- fig_locomo_scaling.pdf
- fig_lora_vs_engram.pdf
- fig_memsystems.pdf
- fig_memsystems_a.pdf
- fig_memsystems_b.pdf
- fig_metric_mismatch.pdf
- fig_mf_finetune.pdf
- fig_multidomain.pdf
- fig_multihop.pdf
- fig_paraphrase.pdf
- fig_pareto.pdf
- fig_pareto_layered.pdf
- fig_pareto_rag.pdf
- fig_pareto_rag.png
- fig_pipeline.pdf
- fig_pretrain_loss.pdf
- fig_rag_scale.pdf
- fig_rag_scale.png
- fig_robustness.pdf
- fig_serving_cdf.pdf
- fig_serving_scale.pdf
- fig_shared_rank.pdf
- fig_strategy_bars.pdf
- fig_user_insertion.pdf
- fig_user_insertion.png
- fig_user_memory_benefits.pdf
- fig_user_memory_benefits.png
- aaai2027.bst
- aaai2027.sty
- abstract.txt
- AnonymousCodeDataAppendix.zip
- AnonymousSubmission2027.pdf
- AnonymousSubmission2027.tex
- build_anonymous_appendix.ps1
- fig_glassbox.py
- fig_landscape.py
- fig_layered_arch.py
- fig_locomo_categories.py
- fig_method_arch.py
- fig_pareto_rag.py
- fig_rag_scale.py
- fig_user_memory_benefits.py
- gen_reorg_figs.py
- LICENSE
- main.tex
- make_all_figures.sh
- make_figures.py
- make_figures_appendix.py
- make_new_figures.py
- README.md
- references.bib
- REPRODUCIBILITY_MISSING_INFORMATION.md
- ReproducibilityChecklist.pdf
- ReproducibilityChecklist.tex
- requirements.txt
- validate_anonymous_appendix.ps1
- insertion_attribution.png
- logitlens_kl.png
- insertion_attribution.png
- logitlens_kl.png
- train_loss.png
- val_bpb.png
- layered_d20_r16_seed_S1.json
- layered_d20_r4_n20.json
- layered_d20_r64_n20.json
- additive_100x100.json
- additive_composition.json
- capacity_ablation_table.txt
- cross_lm_d12_w1280_to_d20.json
- density_layered_d20.json
- depth_d12_1280.json
- depth_d20.json
- engram_d12__factscale_n100.json
- engram_d12__factscale_n1000.json
- engram_d12__factscale_n200.json
- engram_d12__factscale_n500.json
- engram_d12__locomo.json
- engram_d12__scale.json
- engram_d12_large_t05B__locomo.json
- engram_d12_large_t05B__scale.json
- engram_d12_large_t05B__strategies.json
- engram_d12_large_t132B__locomo.json
- engram_d12_large_t132B__scale.json
- engram_d12_large_t132B__strategies.json
- engram_d12_large_t25B__factscale_n100.json
- engram_d12_large_t25B__factscale_n1000.json
- engram_d12_large_t25B__factscale_n200.json
- engram_d12_large_t25B__factscale_n500.json
- engram_d12_large_t25B__locomo.json
- engram_d12_large_t25B__scale.json
- engram_d12_large_t25B__strategies.json
- engram_d12_medium_t05B__locomo.json
- engram_d12_medium_t05B__scale.json
- engram_d12_medium_t05B__strategies.json
- engram_d12_medium_t132B__locomo.json
- engram_d12_medium_t132B__scale.json
- engram_d12_medium_t132B__strategies.json
- engram_d12_medium_t25B__locomo.json
- engram_d12_medium_t25B__scale.json
- engram_d12_medium_t25B__strategies.json
- engram_d12_small_t05B__locomo.json
- engram_d12_small_t05B__scale.json
- engram_d12_small_t05B__strategies.json
- engram_d12_small_t132B__locomo.json
- engram_d12_small_t132B__scale.json
- engram_d12_small_t132B__strategies.json
- engram_d12_small_t25B__locomo.json
- engram_d12_small_t25B__scale.json
- engram_d12_small_t25B__strategies.json
- engram_d12_tiny_t05B__locomo.json
- engram_d12_tiny_t05B__scale.json
- engram_d12_tiny_t05B__strategies.json
- engram_d12_tiny_t132B__locomo.json
- engram_d12_tiny_t132B__scale.json
- engram_d12_tiny_t132B__strategies.json
- engram_d12_tiny_t25B__locomo.json
- engram_d12_tiny_t25B__scale.json
- engram_d12_tiny_t25B__strategies.json
- engram_d12_v2__factscale_n100.json
- engram_d12_v2__factscale_n1000.json
- engram_d12_v2__factscale_n200.json
- engram_d12_v2__factscale_n500.json
- engram_d12_v2__locomo.json
- engram_d12_v2__locomo_cat2_multihop.json
- engram_d12_v2__locomo_cat2_multihop_judge.json
- engram_d12_v2__locomo_cat3_reasoning.json
- engram_d12_v2__locomo_cat3_reasoning_judge.json
- engram_d12_v2__locomo_cat4_opendomain.json
- engram_d12_v2__locomo_cat4_opendomain_judge.json
- engram_d12_v2__locomo_full10.json
- engram_d12_v2__locomo_full10_mt.json
- engram_d12_v2__locomo_full10_mt_judge.json
- engram_d12_v2__locomo_judge.json
- engram_d12_v2__scale.json
- engram_d12_v2__strategies.json
- engram_d12_w1280_optimal__factscale_n100.json
- engram_d12_w1280_optimal__factscale_n1000.json
- engram_d12_w1280_optimal__factscale_n200.json
- engram_d12_w1280_optimal__factscale_n500.json
- engram_d12_w1280_optimal__locomo.json
- engram_d12_w1280_optimal__locomo_cat2_multihop.json
- engram_d12_w1280_optimal__locomo_cat2_multihop_judge.json
- engram_d12_w1280_optimal__locomo_cat3_reasoning.json
- engram_d12_w1280_optimal__locomo_cat3_reasoning_judge.json
- engram_d12_w1280_optimal__locomo_cat4_opendomain.json
- engram_d12_w1280_optimal__locomo_cat4_opendomain_judge.json
- engram_d12_w1280_optimal__locomo_full10.json
- engram_d12_w1280_optimal__locomo_full10_mt.json
- engram_d12_w1280_optimal__locomo_full10_mt_judge.json
- engram_d12_w1280_optimal__locomo_judge.json
- engram_d12_w1280_optimal__locomo_proper.json
- engram_d12_w1280_optimal__locomo_proper_judge.json
- engram_d12_w1280_optimal__memory_systems.json
- engram_d12_w1280_optimal__memory_systems_paraphrase.json
- engram_d12_w1280_optimal__scale.json
- engram_d12_w1280_optimal__strategies.json
- engram_d12_w1280_sft__locomo_full10.json
- engram_d12_w1280_sft__locomo_judge.json
- engram_d12_xlarge_t05B__locomo.json
- engram_d12_xlarge_t05B__scale.json
- engram_d12_xlarge_t05B__strategies.json
- engram_d12_xlarge_t132B__locomo.json
- engram_d12_xlarge_t132B__scale.json
- engram_d12_xlarge_t132B__strategies.json
- engram_d12_xlarge_t25B__locomo.json
- engram_d12_xlarge_t25B__scale.json
- engram_d12_xlarge_t25B__strategies.json
- engram_d20_w1536_optimal__factscale_n100.json
- engram_d20_w1536_optimal__factscale_n1000.json
- engram_d20_w1536_optimal__factscale_n200.json
- engram_d20_w1536_optimal__factscale_n500.json
- engram_d20_w1536_optimal__locomo.json
- engram_d20_w1536_optimal__locomo_cat2_multihop.json
- engram_d20_w1536_optimal__locomo_cat2_multihop_judge.json
- engram_d20_w1536_optimal__locomo_cat3_reasoning.json
- engram_d20_w1536_optimal__locomo_cat3_reasoning_judge.json
- engram_d20_w1536_optimal__locomo_cat4_opendomain.json
- engram_d20_w1536_optimal__locomo_cat4_opendomain_judge.json
- engram_d20_w1536_optimal__locomo_full10.json
- engram_d20_w1536_optimal__locomo_full10_mt.json
- engram_d20_w1536_optimal__locomo_full10_mt_judge.json
- engram_d20_w1536_optimal__locomo_judge.json
- engram_d20_w1536_optimal__locomo_proper.json
- engram_d20_w1536_optimal__locomo_proper_judge.json
- engram_d20_w1536_optimal__memory_systems.json
- engram_d20_w1536_optimal__memory_systems_paraphrase.json
- engram_d20_w1536_optimal__scale.json
- engram_d20_w1536_optimal__strategies.json
- engram_d20_w1536_sft__locomo_full10.json
- engram_d20_w1536_sft__locomo_judge.json
- engram_d8__factscale_n100.json
- engram_d8__factscale_n1000.json
- engram_d8__factscale_n200.json
- engram_d8__factscale_n500.json
- engram_d8__locomo.json
- engram_d8__scale.json
- engram_d8__strategies.json
- engram_d8_large_t05B__locomo.json
- engram_d8_large_t05B__scale.json
- engram_d8_large_t05B__strategies.json
- engram_d8_large_t1B__locomo.json
- engram_d8_large_t1B__scale.json
- engram_d8_large_t1B__strategies.json
- engram_d8_large_t2B__factscale_n100.json
- engram_d8_large_t2B__factscale_n1000.json
- engram_d8_large_t2B__factscale_n200.json
- engram_d8_large_t2B__factscale_n500.json
- engram_d8_large_t2B__locomo.json
- engram_d8_large_t2B__scale.json
- engram_d8_large_t2B__strategies.json
- engram_d8_medium_t05B__locomo.json
- engram_d8_medium_t05B__scale.json
- engram_d8_medium_t05B__strategies.json
- engram_d8_medium_t1B__locomo.json
- engram_d8_medium_t1B__scale.json
- engram_d8_medium_t1B__strategies.json
- engram_d8_medium_t2B__locomo.json
- engram_d8_medium_t2B__scale.json
- engram_d8_medium_t2B__strategies.json
- engram_d8_small_t05B__locomo.json
- engram_d8_small_t05B__scale.json
- engram_d8_small_t05B__strategies.json
- engram_d8_small_t1B__locomo.json
- engram_d8_small_t1B__scale.json
- engram_d8_small_t1B__strategies.json
- engram_d8_small_t2B__locomo.json
- engram_d8_small_t2B__scale.json
- engram_d8_small_t2B__strategies.json
- engram_d8_tiny_t05B__locomo.json
- engram_d8_tiny_t05B__scale.json
- engram_d8_tiny_t05B__strategies.json
- engram_d8_tiny_t1B__locomo.json
- engram_d8_tiny_t1B__scale.json
- engram_d8_tiny_t1B__strategies.json
- engram_d8_tiny_t2B__locomo.json
- engram_d8_tiny_t2B__scale.json
- engram_d8_tiny_t2B__strategies.json
- engram_d8_v1__locomo.json
- engram_d8_v2__factscale_n100.json
- engram_d8_v2__factscale_n1000.json
- engram_d8_v2__factscale_n200.json
- engram_d8_v2__factscale_n500.json
- engram_d8_v2__locomo.json
- engram_d8_v2__locomo_full10.json
- engram_d8_v2__locomo_full10_mt.json
- engram_d8_v2__locomo_full10_mt_judge.json
- engram_d8_v2__locomo_judge.json
- engram_d8_v2__memory_systems.json
- engram_d8_v2__memory_systems_paraphrase.json
- engram_d8_v2__scale.json
- engram_d8_v2__strategies.json
- engram_d8_xlarge_t05B__locomo.json
- engram_d8_xlarge_t05B__scale.json
- engram_d8_xlarge_t05B__strategies.json
- engram_d8_xlarge_t1B__locomo.json
- engram_d8_xlarge_t1B__scale.json
- engram_d8_xlarge_t1B__strategies.json
- engram_d8_xlarge_t2B__locomo.json
- engram_d8_xlarge_t2B__scale.json
- engram_d8_xlarge_t2B__strategies.json
- factscale_table.txt
- glassbox_d12_1280.json
- glassbox_d20.json
- head_to_head_d20.json
- head_to_head_d20_smoke.json
- insertion_opt_d8.json
- insertion_v2_d8.json
- joint_opt_100.json
- joint_opt_1000.json
- joint_opt_30.json
- joint_opt_300.json
- joint_opt_d12_w1280_n100.json
- joint_opt_d12_w1280_n1000.json
- joint_opt_d12_w1280_n1000_s12k.json
- joint_opt_d12_w1280_n2000.json
- joint_opt_d12_w1280_n300.json
- joint_opt_d12_w1280_n500.json
- joint_opt_d20_w1536_n100.json
- joint_opt_d20_w1536_n1000.json
- joint_opt_d20_w1536_n300.json
- joint_opt_mf_d12w1280_n100.json
- joint_opt_mf_d12w1280_n1000.json
- joint_opt_mf_d12w1280_n1000_s12k.json
- joint_opt_mf_d12w1280_n2000.json
- joint_opt_mf_d12w1280_n300.json
- joint_opt_mf_d12w1280_n500.json
- latency_table.csv
- layered_cross_schema_full.json
- layered_d12_w1280_r16.json
- layered_d12_w1280_smoke.json
- layered_d20_r16_dup_partial.json
- layered_d20_r16_full.json
- layered_d20_r4_abl.json
- layered_d20_r64_abl.json
- layered_d20_sft_r16.json
- layered_d20_smoke.json
- layered_d20_trace_r16.json
- layered_d20_trace_v2_r16.json
- layered_judge_cross_schema.json
- layered_judge_cross_schema.predictions.json
- layered_judge_d20.json
- layered_judge_d20.predictions.json
- layered_judge_multi_medical.json
- layered_judge_multi_medical.predictions.json
- layered_judge_opd_personal.json
- layered_judge_opd_personal.predictions.json
- layered_multi_medical.json
- layered_multi_personal.json
- layered_multi_r32_medical.json
- layered_multi_r32_personal.json
- layered_multi_r64_medical.json
- layered_multi_r64_personal.json
- layered_opd_medical.json
- layered_opd_personal.json
- layered_rag_full.json
- layered_rag_scale.json
- layered_rag_scale_smoke.json
- layered_rag_scale_v2.json
- layered_rag_smoke.json
- layered_reverse_xschema.json
- locomo_eval.json
- locomo_layered_d20.json
- longform_gen.json
- longform_gen_d12_w1280.json
- longform_gen_d20_w1536.json
- lora_vs_engram_d12_1280.json
- lora_vs_engram_d20.json
- mechanistic_d12.json
- mechanistic_d8.json
- memory_systems_comparison.json
- memory_systems_paraphrase.json
- multifact_lora_100.json
- multifact_lora_1000.json
- multifact_lora_1000_d12_w1280.json
- multifact_lora_100_d12_w1280.json
- multifact_lora_100_d12_w1280_r128.json
- multifact_lora_100_d12_w1280_r32.json
- multifact_lora_100_d12_w1280_r8.json
- multifact_lora_100_r128.json
- multifact_lora_100_r32.json
- multifact_lora_100_r8.json
- multifact_lora_100_strong.json
- multihop_d20_n63.json
- multihop_probe.json
- multihop_probe_d12_w1280_optimal.json
- multihop_probe_d20_w1536_optimal.json
- multihop_rag.json
- opd_corpus_multi.jsonl
- opt_strong_100.json
- optimal_config.json
- paraphrase_multi.json
- paraphrase_multi_d12.json
- paraphrase_single.json
- paraphrase_single_d12.json
- per_user_table_d12_w1280_100x100.json
- per_user_table_opt_30x100.json
- per_user_table_unembed.json
- per_user_table_xxl_100x100_jointopt.json
- qwen3b_lora_bpb.json
- qwen_rag_full.json
- qwen_rag_scale.json
- qwen_rag_scale_v2.json
- rag_scale_table.csv
- README.md
- replication.json
- scalability_benchmark.json
- scalability_benchmark_xxl_b1.json
- scale_eval_d12.json
- scale_eval_d12_100x100.json
- scale_eval_d12_100x30_salt_strong.json
- scale_eval_d12_1K.json
- scale_eval_d12_e2opt_salt_strong.json
- scale_eval_d8.json
- scale_eval_d8_e2opt_salt.json
- scale_eval_d8_e2opt_salt_strong.json
- scaling_summary.txt
- serving_eval_d12_30u_50f.json
- serving_eval_d12_w1280_100u_100f.json
- serving_eval_d12_w1280_100u_100f_idle.json
- serving_eval_d12_w1280_30u_50f.json
- serving_eval_d8_20u_30f.json
- serving_eval_smoke_d12w1280_5u_10f.json
- sft_baseline_d12.json
- sft_baseline_d12_w1280.json
- t1_2_probe_hash.json
- t1_3_readwrite.json
- t1_4_collisions.json
- user_facts.json
- favicon.svg
- Nav.tsx
- Reveal.tsx
- Scene.tsx
- Hero.tsx
- Resources.tsx
- BrainSplit.tsx
- ContaminationSplit.tsx
- GlassBox.tsx
- KBCrossover.tsx
- LayeredBars.tsx
- MemoryGrid.tsx
- PriceLocator.tsx
- ServingFlow.tsx
- TwoJobs.tsx
- WriteAFact.tsx
- conditions.ts
- facts.ts
- headline.ts
- ragScale.ts
- hash.ts
- hooks.ts
- App.tsx
- index.css
- main.tsx
- theme.ts
- vite-env.d.ts
- .gitignore
- eslint.config.js
- index.html
- package-lock.json
- package.json
- postcss.config.js
- README.md
- tailwind.config.js
- tsconfig.app.json
- tsconfig.json
- tsconfig.node.json
- vite.config.ts
- .gitignore
- LICENSE
- NOTICE
- README.md
- REPRODUCE.md
# CDN으로 사용하기
jsDelivrjsDelivr는 공개 GitHub 리포지토리를 별도 설정 없이 CDN으로 즉시 서빙합니다. 버전과 파일을 고르면 웹페이지에 바로 붙일 수 있는 링크와 예시 코드가 만들어집니다.
명령어 용어집
이 문서에서 사용된 명령어를 모아봤습니다. 낯선 명령어가 있다면 펼쳐서 확인해보세요.
npm install
설명 보기 ▼
npm install
Install Node packages.
npm {{[i|install]}}
Install dependencies listed in `package.json`:
npm {{[i|install]}} {{package_name}}@{{version}}
Download a specific version of a package and add it to the list of dependencies in `package.json`:
npm {{[i|install]}} {{package_name}} {{[-D|--save-dev]}}
Download the latest version of a package and add it to the list of dev dependencies in `package.json`:
npm run
설명 보기 ▼
npm run
Run a script.
npm run
List available scripts:
npm run {{script_name}}
Run a script:
npm run {{script_name}} -- {{argument}} {{--option}}
Pass arguments to a script:
pip install
설명 보기 ▼
pip install
Install Python packages.
pip install {{package1 package2 ...}}
Install one or more packages:
pip install {{package1 package2 ...}} {{[-U|--upgrade]}}
Upgrade all specified packages to the latest version, installing any that are not already present:
pip install {{package}}=={{version}}
Install a specific version of a package:
