{"id":"W4409601530","doi":"10.1007/s00521-025-11164-y","title":"Self-evaluation of LLMs on challenging LLM-generated STEM MCQs","year":2025,"lang":"en","type":"article","venue":"Neural Computing and Applications","topic":"Advanced Data Storage Technologies","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":false,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"","keywords":"Computational Science and Engineering; Computer science; Mathematics education; Psychology; Machine learning","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":[0.001999269,0.0005918039,0.0005154727,0.0008863773,0.0004452358,0.0009171186,0.0009987915,0.001133294,0.009157989],"category_scores_gemma":[0.01442723,0.0001135105,0.0002975258,0.0004774787,0.0003945258,0.0007250856,0.001256692,0.0005388904,0.001997451],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006834164,"about_ca_system_score_gemma":0.0007149146,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003519274,"about_ca_topic_score_gemma":0.005220765,"domain_scores_codex":[0.9990423,0.000276883,0.00004446861,0.0001780428,0.0003762716,0.00008201892],"domain_scores_gemma":[0.9944225,0.002718281,0.000181094,0.0005920023,0.001796749,0.0002892809],"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":[0.002942131,0.0009477813,0.01552092,0.0009916363,0.0002022472,0.0004897126,0.0008149744,0.4011263,0.0522203,0.01018067,0.0423724,0.4721909],"study_design_scores_gemma":[0.0001018192,0.0006274286,0.005377158,0.00006033326,0.00003249766,0.0001148536,0.0004025594,0.9559014,0.02361446,0.003183444,0.01054926,0.00003476729],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8305303,0.0009258314,0.1083264,0.000705529,0.0006493807,0.0004504502,0.002689243,0.007454679,0.04826829],"genre_scores_gemma":[0.9371846,0.00009385331,0.04869304,0.000180941,0.00005037941,0.0001583824,0.004194525,0.0004365704,0.009007732],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.009157989,"threshold_uncertainty_score":0.03063649,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03904450454822871,"score_gpt":0.3134778578845829,"score_spread":0.2744333533363542,"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."}}