{"id":"W4414158748","doi":"10.1007/978-981-95-0988-1_1","title":"Evaluating the Behavior of Small Language Models in Answering Binary Questions","year":2025,"lang":"en","type":"book-chapter","venue":"Communications in computer and information science","topic":"Topic Modeling","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":false,"ca_institutions":"McGill University; Mila - Quebec Artificial Intelligence Institute","funders":"","keywords":"Security token; Binary number; Natural language; Language model; Binary classification; Natural (archaeology)","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":[0.001622195,0.0001303682,0.000177989,0.0008371614,0.0002411233,0.0001995796,0.003195525,0.00007515846,0.000001357623],"category_scores_gemma":[0.00005263889,0.0001155259,0.00003288897,0.0004675116,0.0003636232,0.00280168,0.00258132,0.000385751,0.000001793366],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001035184,"about_ca_system_score_gemma":0.0003626404,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00009710401,"about_ca_topic_score_gemma":0.00004895793,"domain_scores_codex":[0.9986287,0.00004810797,0.0006759714,0.0002005603,0.0002955561,0.0001511144],"domain_scores_gemma":[0.9971955,0.0002682541,0.0002449834,0.002032159,0.0002280535,0.00003100111],"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":[6.065017e-7,0.00001097006,0.00003273433,0.00002193613,0.000001335144,2.207044e-7,0.003847156,0.02315914,0.00001479532,0.6038916,0.000003465237,0.3690161],"study_design_scores_gemma":[0.000113388,0.00002441008,0.0005760434,0.0003383862,0.000004024061,0.000004429651,0.00004600398,0.9929265,0.000005831434,0.005359339,0.0004930256,0.0001086472],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.002292651,0.0009393343,0.9074351,0.0006805739,0.0002318634,0.0007633555,0.000008524815,0.00006063872,0.087588],"genre_scores_gemma":[0.2736231,0.001280291,0.7232351,0.0005335609,0.00002069297,0.0001774823,0.00002046927,0.000007489718,0.001101767],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.9697673,"threshold_uncertainty_score":0.5938134,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1177777415211075,"score_gpt":0.3699668967910038,"score_spread":0.2521891552698963,"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."}}