{"id":"W4390041668","doi":"10.48550/arxiv.2312.11690","title":"Agent-based Learning of Materials Datasets from Scientific Literature","year":2023,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Machine Learning in Materials Science","field":"Materials Science","cited_by":12,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada; University of Toronto","keywords":"Computer science; Scalability; Information extraction; Data science; Bottleneck; Consistency (knowledge bases); Artificial intelligence; Resource (disambiguation); Workflow; Natural language; The Internet; Machine learning; World Wide Web; Database","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","insufficient_payload"],"consensus_categories":["insufficient_payload"],"category_scores_codex":[0.002047134,0.000501758,0.0007403335,0.0004989346,0.0004184167,0.0009512338,0.002430565,0.0004337306,0.001935887],"category_scores_gemma":[0.000584316,0.0005339886,0.0001753212,0.0009107868,0.0006745416,0.0003640772,0.00237865,0.0006220393,0.001555505],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001390008,"about_ca_system_score_gemma":0.000325039,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0009087436,"about_ca_topic_score_gemma":0.00004225861,"domain_scores_codex":[0.9954993,0.0008915413,0.000562125,0.002070078,0.0003686652,0.0006083415],"domain_scores_gemma":[0.9964368,0.0003000142,0.0009927341,0.001822236,0.0002490599,0.0001991686],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0001037704,0.00005757907,0.00135082,0.0003253295,0.00001846782,0.0002490502,0.0002326618,0.4575538,0.5378834,0.0009973971,0.001216314,0.00001137986],"study_design_scores_gemma":[0.001465013,0.0001822425,0.006999507,0.002872134,0.0003650473,0.000004343363,0.0001612913,0.1597334,0.8060787,0.01554605,0.004569505,0.002022773],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9769003,0.00005634537,0.01131985,0.00004906388,0.004941626,0.0003603228,0.005832824,0.0004705726,0.00006910293],"genre_scores_gemma":[0.9910465,0.00002431286,0.002579999,0.00003004385,0.000167885,0.000002767575,0.004062349,0.00006248929,0.002023604],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2978204,"threshold_uncertainty_score":0.9997112,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06760741861473786,"score_gpt":0.218393963363026,"score_spread":0.1507865447482881,"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."}}