{"id":"W4320710261","doi":"10.1080/08989621.2023.2179919","title":"Can ChatGPT be trusted to provide reliable estimates?","year":2023,"lang":"en","type":"letter","venue":"Accountability in Research","topic":"Explainable Artificial Intelligence (XAI)","field":"Computer Science","cited_by":21,"is_retracted":false,"has_abstract":true,"ca_institutions":"Thompson Rivers University","funders":"","keywords":"Direct Anonymous Attestation; Computer security; Internet privacy; Computer science; Scientific misconduct; Trusted Computing; Medicine","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[{"model":"gemma","categories":["metaresearch"],"domain":"methods","study_design":"not_applicable","genre":"commentary","about_ca_system":false,"about_ca_topic":false,"confidence":"low","status":"direct model label, unvalidated"},{"model":"gpt","categories":[],"domain":null,"study_design":"not_applicable","genre":"commentary","about_ca_system":false,"about_ca_topic":false,"confidence":"low","status":"direct model label, unvalidated"}],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","scholarly_communication","open_science","research_integrity","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.009342914,0.0005273293,0.0007534989,0.001858433,0.0004410962,0.001384308,0.006417305,0.001055164,0.0001845705],"category_scores_gemma":[0.006289001,0.0005359224,0.0001458903,0.006809411,0.0004091356,0.0009048234,0.003117937,0.00566786,0.002752776],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002155269,"about_ca_system_score_gemma":0.001586952,"about_ca_topic_candidate":true,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0552473,"about_ca_topic_score_gemma":0.01467286,"domain_scores_codex":[0.9897093,0.0007376699,0.001039162,0.002378065,0.00324031,0.002895544],"domain_scores_gemma":[0.990083,0.003800906,0.0001268163,0.004290033,0.001413774,0.0002855085],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00002738438,0.00009874188,0.001560197,0.0004169177,0.00001639974,0.000709967,0.00288145,0.0003040211,0.0001124626,0.0007451918,0.9902774,0.002849872],"study_design_scores_gemma":[0.0001452054,0.0003534491,0.0008600486,0.0004085524,0.000005800862,0.00001395985,0.0005753678,0.01545185,0.004539295,0.0370457,0.9396437,0.0009570711],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"commentary","genre_gemma":"commentary","genre_scores_codex":[0.01162209,0.00003833435,0.0007755681,0.9808806,0.0009358461,0.00294943,0.00005888327,0.0005706608,0.002168605],"genre_scores_gemma":[0.04022966,0.0001115583,0.0143819,0.9107402,0.003954655,0.006423162,0.0003753536,0.0004902538,0.02329328],"genre_candidate":"commentary","genre_consensus":"commentary","teacher_disagreement_score":0.0701404,"threshold_uncertainty_score":0.9997092,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2534849222970801,"score_gpt":0.4491510783357754,"score_spread":0.1956661560386953,"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."}}