{"id":"W6910119171","doi":"10.48448/g98m-z359","title":"Do Large Language Models Know How Much They Know?","year":2024,"lang":"en","type":"other","venue":"Underline Science Inc.","topic":"","field":"","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal; Polytechnique Montréal","funders":"","keywords":"Scope (computer science); Software deployment; Benchmark (surveying); Language model; Property (philosophy); Key (lock)","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":[0.006776613,0.001006005,0.0009455884,0.001667389,0.0008556254,0.006348808,0.00181188,0.002034338,0.007705277],"category_scores_gemma":[0.0687736,0.0006146088,0.0006286354,0.001458861,0.002353796,0.02154474,0.002134104,0.002907671,0.003147279],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001682549,"about_ca_system_score_gemma":0.002154356,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005955291,"about_ca_topic_score_gemma":0.00618321,"domain_scores_codex":[0.996207,0.001951447,0.0001480089,0.0008601922,0.0006335293,0.0001997706],"domain_scores_gemma":[0.9493138,0.03636865,0.003156901,0.007067697,0.002844689,0.001248371],"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":[0.001242805,0.0004292504,0.0741148,0.001140898,0.0005589474,0.0004205394,0.006296275,0.09896436,0.00581464,0.3102677,0.03723666,0.4635131],"study_design_scores_gemma":[0.00006337281,0.0001504261,0.01030017,0.0003872561,0.000143203,0.0004071895,0.002272921,0.3074264,0.003943618,0.6382694,0.03648303,0.0001530139],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3806484,0.01384755,0.4019591,0.0589841,0.0006457599,0.0002179532,0.005755997,0.005222845,0.1327184],"genre_scores_gemma":[0.9427626,0.003422719,0.04054734,0.00213478,0.0003342622,0.0001392765,0.002937889,0.0007303359,0.006990918],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007705277,"threshold_uncertainty_score":0.0358386,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02435099445772429,"score_gpt":0.3133688716432343,"score_spread":0.28901787718551,"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."}}