{"id":"W3017343191","doi":"10.48550/arxiv.2004.07213","title":"Toward Trustworthy AI Development: Mechanisms for Supporting Verifiable Claims","year":2020,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":125,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University; Université du Québec à Montréal; Polytechnique Montréal","funders":"","keywords":"Verifiable secret sharing; Trustworthiness; Computer science; Development (topology); Computer security; Data science; Programming language; Mathematics","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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0005925567,0.0005234152,0.0006144872,0.0002495059,0.0004451484,0.000271219,0.003049958,0.0004354843,0.00006011601],"category_scores_gemma":[0.000172247,0.0006547449,0.0003146229,0.0006470673,0.00006230196,0.0006205154,0.003992315,0.001156253,0.00008524468],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000354435,"about_ca_system_score_gemma":0.000845005,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00004005732,"about_ca_topic_score_gemma":0.000007507799,"domain_scores_codex":[0.9966071,0.0001397373,0.0004492269,0.001854569,0.0001890256,0.00076034],"domain_scores_gemma":[0.9977316,0.0001702179,0.0005851116,0.0009851514,0.000219612,0.0003082642],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00005839538,0.00003916335,0.0002982081,0.0002869227,0.000151982,0.0002983655,0.0009747109,0.5179328,0.00004985514,0.4768997,0.0003476516,0.002662233],"study_design_scores_gemma":[0.0006638207,0.00005524793,0.00006361349,0.00008718817,0.00006416435,0.000004133502,0.00011404,0.8056325,0.0005964079,0.188808,0.003213908,0.0006970441],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.006679377,0.00001520041,0.9890161,0.0006999933,0.001334929,0.0006696719,0.00001067537,0.0006829396,0.0008911107],"genre_scores_gemma":[0.8352531,0.00001048819,0.1632858,0.0004528632,0.0001452126,0.00000636191,0.00006176306,0.00005031781,0.0007340682],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8285738,"threshold_uncertainty_score":0.9995904,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09684774672380762,"score_gpt":0.2238924999907693,"score_spread":0.1270447532669616,"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."}}