{"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/wilson","entry":"wilson","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":5,"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":10,"n_samples_ran":8,"n_samples_fingerprinted":5,"n_places":10,"n_places_pointer_only":7,"by_status":{"ran_honours":0,"ran_violates":0,"ran_draft_wrong":0,"ran_fixture":0,"ran":8,"unverified":2},"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":"2609.01345","paper":"/paper/arxiv-2609-01345","title":"Cheap Verifiers, Large Blind Spots: Measuring the Reliability Cost of Cost-Saving Cascades","date":null,"month_inferred_from_arxiv_id":"2026-09","title_source":"syntology","repo":"AltSlate-Labs/cascade-blindspot","path":"experiments/phase0_real/make_real_figures.py","file_url":"https://github.com/AltSlate-Labs/cascade-blindspot/blob/HEAD/experiments/phase0_real/make_real_figures.py","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"code_sha256_prefix":"eae8880c42a8f847","mcp_get_code":{"code_sha256":"eae8880c42a8f847"}},{"arxiv_id":"2608.23640","paper":"/paper/arxiv-2608-23640","title":"Auditing the Synthetic Memoir: Measuring Scene-Level Confabulation in LLM-Generated Autobiography Against the Documented Record of the Life It Describes","date":null,"month_inferred_from_arxiv_id":"2026-08","title_source":"syntology","repo":"heathriel/synthetic-memoir-audit","path":"analysis/02_analyze.py","file_url":"https://github.com/heathriel/synthetic-memoir-audit/blob/HEAD/analysis/02_analyze.py","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"code_sha256_prefix":"364c89fa74185151","mcp_get_code":{"code_sha256":"364c89fa74185151"}},{"arxiv_id":"2608.23640","paper":"/paper/arxiv-2608-23640","title":"Auditing the Synthetic Memoir: Measuring Scene-Level Confabulation in LLM-Generated Autobiography Against the Documented Record of the Life It Describes","date":null,"month_inferred_from_arxiv_id":"2026-08","title_source":"syntology","repo":"heathriel/synthetic-memoir-audit","path":"analysis/02_analyze_public.py","file_url":"https://github.com/heathriel/synthetic-memoir-audit/blob/HEAD/analysis/02_analyze_public.py","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"code_sha256_prefix":"5f0ab59efce198c6","mcp_get_code":{"code_sha256":"5f0ab59efce198c6"}},{"arxiv_id":"2608.23640","paper":"/paper/arxiv-2608-23640","title":"Auditing the Synthetic Memoir: Measuring Scene-Level Confabulation in LLM-Generated Autobiography Against the Documented Record of the Life It Describes","date":null,"month_inferred_from_arxiv_id":"2026-08","title_source":"syntology","repo":"heathriel/synthetic-memoir-audit","path":"analysis/07_blind_replication.py","file_url":"https://github.com/heathriel/synthetic-memoir-audit/blob/HEAD/analysis/07_blind_replication.py","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"code_sha256_prefix":"ea1bc45efe333e30","mcp_get_code":{"code_sha256":"ea1bc45efe333e30"}},{"arxiv_id":"2608.23640","paper":"/paper/arxiv-2608-23640","title":"Auditing the Synthetic Memoir: Measuring Scene-Level Confabulation in LLM-Generated Autobiography Against the Documented Record of the Life It Describes","date":null,"month_inferred_from_arxiv_id":"2026-08","title_source":"syntology","repo":"heathriel/synthetic-memoir-audit","path":"analysis/08_replication_stats.py","file_url":"https://github.com/heathriel/synthetic-memoir-audit/blob/HEAD/analysis/08_replication_stats.py","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"code_sha256_prefix":"05188b4d85e04c82","mcp_get_code":{"code_sha256":"05188b4d85e04c82"}},{"arxiv_id":"2608.08212","paper":"/paper/arxiv-2608-08212","title":"Harmful Content Is Not Enough: Continuation Framing Moderates In-Context Emergent Misalignment","date":null,"month_inferred_from_arxiv_id":"2026-08","title_source":"syntology","repo":"PeiYangLiu/icl-em-format-control","path":"run_format_ladder.py","file_url":"https://github.com/PeiYangLiu/icl-em-format-control/blob/HEAD/run_format_ladder.py","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"code_sha256_prefix":"269cc0daaa143874","mcp_get_code":{"code_sha256":"269cc0daaa143874"}},{"arxiv_id":"2607.24165","paper":"/paper/arxiv-2607-24165","title":"Do Current Retrievers Cover All the Evidence? A Controlled Study of Conjunctive Cross-Page Retrieval","date":null,"month_inferred_from_arxiv_id":"2026-07","title_source":"syntology","repo":"sunggukcha/n_clue","path":"audit/admin/score_audit.py","file_url":"https://github.com/sunggukcha/n_clue/blob/HEAD/audit/admin/score_audit.py","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"code_sha256_prefix":"5a97074bff6ce952","mcp_get_code":{"code_sha256":"5a97074bff6ce952"}},{"arxiv_id":"2605.07093","paper":"/paper/arxiv-2605-07093","title":"The Translation Tax Is Not a Scalar: A Counterfactual Audit of English-Source Cue Inheritance in Chinese Multilingual Benchmarks","date":null,"month_inferred_from_arxiv_id":"2026-05","title_source":"syntology","repo":"chi-mi-rvard/translation-tax-supplement","path":"run_e3_comparison.py","file_url":"https://github.com/chi-mi-rvard/translation-tax-supplement/blob/HEAD/run_e3_comparison.py","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NOASSERTION","inline_ok":false,"code_sha256_prefix":"0ee1631d42dfb3db","mcp_get_code":{"code_sha256":"0ee1631d42dfb3db"}},{"arxiv_id":"2604.26052","paper":"/paper/arxiv-2604-26052","title":"From Prompt Risk to Response Risk: Paired Analysis of Safety Behavior of Large Language Models","date":null,"month_inferred_from_arxiv_id":"2026-04","title_source":"syntology","repo":"microsoft/PairedSafety","path":"analysis/public_mini_eval/04_analyze.py","file_url":"https://github.com/microsoft/PairedSafety/blob/HEAD/analysis/public_mini_eval/04_analyze.py","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"code_sha256_prefix":"e96ea8226b0e733b","mcp_get_code":{"code_sha256":"e96ea8226b0e733b"}},{"arxiv_id":"2004.07999","paper":"/paper/vibe-a-tool-for-measuring-and-mitigating-bias","title":"REVISE: A Tool for Measuring and Mitigating Bias in Visual Datasets","date":"2020-04-16","month_inferred_from_arxiv_id":null,"title_source":"archive","repo":"princetonvisualai/revise-tool","path":"measurements/prerun_analyzegeo.py","file_url":"https://github.com/princetonvisualai/revise-tool/blob/HEAD/measurements/prerun_analyzegeo.py","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"code_sha256_prefix":"bed51f23c7b162eb","mcp_get_code":{"code_sha256":"bed51f23c7b162eb"}}]}