{"id":"W4289523162","doi":"10.1073/pnas.2123433119","title":"A neural network solves, explains, and generates university math problems by program synthesis and few-shot learning at human level","year":2022,"lang":"en","type":"article","venue":"Proceedings of the National Academy of Sciences","topic":"Topic Modeling","field":"Computer Science","cited_by":182,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Artificial neural network; Benchmark (surveying); Linear algebra; Code (set theory); Artificial intelligence; Algorithm; Algebra over a field; Machine learning; Theoretical computer science; Mathematics; Programming language; Pure mathematics; Set (abstract data type)","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001012105,0.002006387,0.0005357964,0.0008561527,0.0004210018,0.00109799,0.002329578,0.00190584,0.006858732],"category_scores_gemma":[0.006624418,0.0004961293,0.001253008,0.0005922497,0.0005954323,0.001848836,0.0009742281,0.002472691,0.003405978],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001638582,"about_ca_system_score_gemma":0.001572212,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01147883,"about_ca_topic_score_gemma":0.02256998,"domain_scores_codex":[0.9991413,0.0001581123,0.00003778478,0.00046824,0.0001239145,0.00007065983],"domain_scores_gemma":[0.9979782,0.001284691,0.00008354671,0.0003043874,0.0002585765,0.00009078075],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0005678234,0.0008405251,0.008509059,0.001084634,0.0002968543,0.0003444401,0.0004225289,0.1856262,0.02126575,0.004707247,0.05258769,0.7237472],"study_design_scores_gemma":[0.00007369665,0.0001649431,0.001323045,0.0000486983,0.00004694534,0.00008822395,0.00008222945,0.9752725,0.01046904,0.005634849,0.006773788,0.00002181646],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3380747,0.003912516,0.5180833,0.0031582,0.001042981,0.0008935091,0.009730894,0.09614225,0.02896167],"genre_scores_gemma":[0.6165234,0.000659593,0.3393121,0.002197467,0.0001420464,0.000716312,0.02514357,0.001443535,0.01386199],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01147883,"threshold_uncertainty_score":0.02294475,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07421882056051196,"score_gpt":0.269741645150327,"score_spread":0.195522824589815,"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."}}