{"id":"W4406983107","doi":"10.1073/pnas.2416866122","title":"Errors are robustly tamed in cumulative knowledge processes","year":2025,"lang":"en","type":"article","venue":"Proceedings of the National Academy of Sciences","topic":"Logic, Reasoning, and Knowledge","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Toronto Metropolitan University","funders":"Army Research Office; Office of Naval Research; Multidisciplinary University Research Initiative; National Science Foundation","keywords":"Heuristics; Computer science; Simple (philosophy); Probabilistic logic; Adversarial system; Bounded function; Fraction (chemistry); Process (computing); Theoretical computer science; Artificial intelligence; Mathematics; Epistemology; Programming language","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001342628,0.0001224505,0.0002146174,0.000415854,0.0001978138,0.00005614641,0.002321506,0.00009166474,0.000003231592],"category_scores_gemma":[0.002112994,0.00008053079,0.00006287362,0.003932139,0.0006457012,0.0009201878,0.0004840792,0.0001945441,0.000002296288],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006750962,"about_ca_system_score_gemma":0.0002153577,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000005401416,"about_ca_topic_score_gemma":0.000001685755,"domain_scores_codex":[0.9983536,0.00001293766,0.0003635291,0.0003828942,0.0006715678,0.0002154843],"domain_scores_gemma":[0.9985297,0.0002905371,0.0004348751,0.0000145107,0.0006993576,0.0000310757],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00001186844,0.0002641524,0.06563912,0.0004323493,0.00002066639,2.267018e-8,0.003326327,0.0005244585,0.005833071,0.919875,0.002622203,0.00145078],"study_design_scores_gemma":[0.0005777643,0.00006750934,0.3573228,0.001029884,0.00001259909,0.000004961353,0.0008470649,0.03247057,0.1400305,0.4661256,0.001232311,0.0002784275],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7689178,0.002062104,0.0003281379,0.008933958,0.0001708827,0.0006308969,0.000006416629,0.00009369979,0.2188561],"genre_scores_gemma":[0.9958117,0.00002695849,0.003039004,0.0001760504,0.00003052104,0.00001785234,2.831818e-8,0.000002046629,0.0008958259],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.4537494,"threshold_uncertainty_score":0.4313975,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05174160605511081,"score_gpt":0.3288787505353769,"score_spread":0.2771371444802661,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}