{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/code/extract-question","entry":"extract_question","source":"Syntology graph, per-sample; not an archive number","read_at":"2026-09-24T18:15:14+00:00","claim":"Names are grouped by exact entry-name string. Same-named routines are NOT asserted to be equivalent; 'ran' means executed on a synthesized fixture, not correctness. n_samples_ran = sum of by_status over every status except 'unverified' (ran_draft_wrong and ran_fixture are failures of Syntology's instrument, not of the code); n_papers_ran = papers with at least one such sample.","status_vocabulary":{"ran_honours":"ran, honoured the contract we drafted","ran_violates":"ran, violated the contract we drafted","ran_draft_wrong":"ran; our contract draft was wrong, not the code","ran_fixture":"ran; our fixture could not drive it","ran":"ran on a synthesized input","unverified":"unverified (harvested, no recorded run)"},"n_papers":7,"n_papers_ran":3,"units":"n_samples, n_samples_ran, n_samples_fingerprinted and by_status count distinct code bodies (code_sha256); n_places and n_places_pointer_only count places, one per (paper, code body) pair, which is also the unit of the samples list","n_samples":7,"n_samples_ran":3,"n_samples_fingerprinted":1,"n_places":7,"n_places_pointer_only":2,"by_status":{"ran_honours":0,"ran_violates":0,"ran_draft_wrong":2,"ran_fixture":0,"ran":1,"unverified":4},"syntology":{"atlas_url":null,"mcp":null,"mcp_per_sample":{"tool":"get_code","arguments_in":"samples[].mcp_get_code"},"developers":"https://syntology.ai/developers"},"samples":[{"arxiv_id":"2604.25444","paper":"/paper/arxiv-2604-25444","title":"One Refiner to Unlock Them All: Inference-Time Reasoning Elicitation via Reinforcement Query Refinement","date":null,"month_inferred_from_arxiv_id":"2026-04","title_source":"syntology","repo":"newera-xiao/ReQueR","path":"verl/rephrase_cold_start_sft/generate_sft_data.py","file_url":"https://github.com/newera-xiao/ReQueR/blob/HEAD/verl/rephrase_cold_start_sft/generate_sft_data.py","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"code_sha256_prefix":"832a35aab7fa499f","mcp_get_code":{"code_sha256":"832a35aab7fa499f"}},{"arxiv_id":"2510.11520","paper":"/paper/arxiv-2510-11520","title":"mmWalk: Towards Multi-modal Multi-view Walking Assistance","date":null,"month_inferred_from_arxiv_id":"2025-10","title_source":"syntology","repo":"KediYing/mmWalk","path":"eval_gpt.py","file_url":"https://github.com/KediYing/mmWalk/blob/HEAD/eval_gpt.py","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"code_sha256_prefix":"841d4c776de0854a","mcp_get_code":{"code_sha256":"841d4c776de0854a"}},{"arxiv_id":"2506.12860","paper":"/paper/qfft-question-free-fine-tuning-for-adaptive","title":"QFFT, Question-Free Fine-Tuning for Adaptive Reasoning","date":"2025-06-15","month_inferred_from_arxiv_id":null,"title_source":"archive","repo":"lwl-cpu/question-free-fine-tuning","path":"eval/utils/analysis_difficulty.py","file_url":"https://github.com/lwl-cpu/question-free-fine-tuning/blob/HEAD/eval/utils/analysis_difficulty.py","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"code_sha256_prefix":"3bf1dc9721e95221","mcp_get_code":{"code_sha256":"3bf1dc9721e95221"}},{"arxiv_id":"2505.18673","paper":"/paper/cross-lingual-pitfalls-automatic-probing","title":"Cross-Lingual Pitfalls: Automatic Probing Cross-Lingual Weakness of Multilingual Large Language Models","date":"2025-05-24","month_inferred_from_arxiv_id":null,"title_source":"archive","repo":"xzx34/Cross-Lingual-Pitfalls","path":"utils/prompt.py","file_url":"https://github.com/xzx34/Cross-Lingual-Pitfalls/blob/HEAD/utils/prompt.py","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"code_sha256_prefix":"99fb32bf3b96639a","mcp_get_code":{"code_sha256":"99fb32bf3b96639a"}},{"arxiv_id":"2412.00535","paper":"/paper/fullstack-bench-evaluating-llms-as-full-stack","title":"FullStack Bench: Evaluating LLMs as Full Stack Coders","date":"2024-11-30","month_inferred_from_arxiv_id":null,"title_source":"archive","repo":"bytedance/sandboxfusion","path":"sandbox/datasets/live_code_bench.py","file_url":"https://github.com/bytedance/sandboxfusion/blob/HEAD/sandbox/datasets/live_code_bench.py","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"code_sha256_prefix":"cb2a2f189c014acf","mcp_get_code":{"code_sha256":"cb2a2f189c014acf"}},{"arxiv_id":"2409.02389","paper":"/paper/multi-modal-situated-reasoning-in-3d-scenes","title":"Multi-modal Situated Reasoning in 3D Scenes","date":"2024-09-04","month_inferred_from_arxiv_id":null,"title_source":"archive","repo":"MSR3D/MSR3D","path":"evaluator/evaluate_msqa.py","file_url":"https://github.com/MSR3D/MSR3D/blob/HEAD/evaluator/evaluate_msqa.py","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"Apache-2.0","inline_ok":true,"code_sha256_prefix":"88bd34c3dd6b90a7","mcp_get_code":{"code_sha256":"88bd34c3dd6b90a7"}},{"arxiv_id":"2310.00305","paper":"/paper/towards-llm-based-fact-verification-on-news","title":"Towards LLM-based Fact Verification on News Claims with a Hierarchical Step-by-Step Prompting Method","date":"2023-09-30","month_inferred_from_arxiv_id":null,"title_source":"archive","repo":"jadecurl/hiss","path":"HiSS.py","file_url":"https://github.com/jadecurl/hiss/blob/HEAD/HiSS.py","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"code_sha256_prefix":"ebc2fb9ab39dd72e","mcp_get_code":{"code_sha256":"ebc2fb9ab39dd72e"}}]}