{"id":"W4386942346","doi":"10.1145/3624918.3625336","title":"Retrieving Supporting Evidence for Generative Question Answering","year":2023,"lang":"en","type":"preprint","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":59,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Question answering; Computer science; Statement (logic); Hallucinating; Information retrieval; Pipeline (software); Natural language processing; Questions and answers; Artificial intelligence; Domain (mathematical analysis); Linguistics; 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":[0.001823782,0.0002225615,0.0002754785,0.000146621,0.0001502739,0.0004757181,0.001131821,0.0001945266,0.000005512115],"category_scores_gemma":[0.001540236,0.0002249245,0.0001370053,0.000176259,0.00001361639,0.0004871921,0.002045964,0.0003600785,0.0000168892],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001665362,"about_ca_system_score_gemma":0.0002497205,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001941365,"about_ca_topic_score_gemma":0.00004478819,"domain_scores_codex":[0.9976858,0.00007240353,0.0005384107,0.001020606,0.0003006725,0.0003821004],"domain_scores_gemma":[0.9979923,0.0004993667,0.0003055025,0.0009001088,0.0002257302,0.00007692585],"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":[0.00002500211,0.00004241812,0.004309095,0.002442798,0.0001839513,0.00007639434,0.008902353,0.3508789,0.009713285,0.3406356,0.003628656,0.2791615],"study_design_scores_gemma":[0.00005492533,0.00002107804,0.0002253036,0.0007652372,0.000008793954,0.000002264676,0.00002979932,0.9684944,0.003800265,0.02622097,0.0001012172,0.0002757536],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.006197507,0.0001986604,0.9872577,0.002538455,0.002112708,0.0006079085,0.00000240701,0.0009252094,0.0001594529],"genre_scores_gemma":[0.2278222,0.00006186822,0.769425,0.0001791751,0.00055561,0.0001576342,0.000008410245,0.00002750387,0.001762541],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.6176155,"threshold_uncertainty_score":0.9172153,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2230135931799158,"score_gpt":0.4047392601166221,"score_spread":0.1817256669367063,"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."}}