{"id":"W4389520703","doi":"10.18653/v1/2023.emnlp-main.489","title":"TheoremQA: A Theorem-driven Question Answering Dataset","year":2023,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":30,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Benchmark (surveying); Computer science; Domain (mathematical analysis); Code (set theory); Question answering; Quality (philosophy); Baseline (sea); Artificial intelligence; Programming language; Mathematics; Epistemology","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":[0.000398782,0.00007147629,0.00006941896,0.00008497613,0.00006908649,0.00009730781,0.0006271822,0.00002930559,0.00003068213],"category_scores_gemma":[0.00003852783,0.00005881452,0.0000210481,0.0003174606,0.00001794473,0.0003667557,0.0003646357,0.00006572349,0.0004859669],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001580161,"about_ca_system_score_gemma":0.00001867329,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00004238377,"about_ca_topic_score_gemma":0.0000114727,"domain_scores_codex":[0.9992,0.00004679355,0.0001143983,0.0002790095,0.0001593326,0.0002004524],"domain_scores_gemma":[0.9992439,0.00005792758,0.00002265683,0.0006090511,0.00001807522,0.00004836512],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[8.418002e-7,0.000005398758,0.0001653698,0.000004281057,0.00000396976,0.00001415004,0.0001977884,0.002184302,0.0005869007,0.9651733,0.004447564,0.02721613],"study_design_scores_gemma":[0.0001018605,0.00001452342,0.0005738896,0.00001309044,0.000001656078,0.000007672766,0.00002822923,0.9392995,0.0006243896,0.04843084,0.01078679,0.0001175739],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.009501349,0.000008874522,0.9829026,0.001542268,0.0002782053,0.00008104541,0.00001709498,0.0007606241,0.004907866],"genre_scores_gemma":[0.9300612,0.00001529977,0.0686847,0.0004763951,0.0001122472,0.00001331488,0.0001233931,0.000009414991,0.0005040065],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9371152,"threshold_uncertainty_score":0.6246284,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0263743278939783,"score_gpt":0.2841081159769848,"score_spread":0.2577337880830065,"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."}}