{"id":"W4417004307","doi":"10.48550/arxiv.2512.02304","title":"When Does Verification Pay Off? A Closer Look at LLMs as Solution Verifiers","year":2025,"lang":"","type":"preprint","venue":"ArXiv.org","topic":"Machine Learning in Materials Science","field":"Materials Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Institute for Information and Communications Technology Promotion; Natural Sciences and Engineering Research Council of Canada; Ministry of Science and ICT, South Korea","keywords":"Metric (unit); Domain (mathematical analysis); Base (topology); Solver; Scale (ratio); Problem solver; Logical framework","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.02788303,0.0008143864,0.001352567,0.001661549,0.001157463,0.005093123,0.002440352,0.002736336,0.005511087],"category_scores_gemma":[0.1535148,0.0007222456,0.0009951455,0.001350475,0.003936931,0.01739689,0.005066604,0.005609637,0.001484179],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002420027,"about_ca_system_score_gemma":0.002724534,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002164031,"about_ca_topic_score_gemma":0.002358858,"domain_scores_codex":[0.9784585,0.01256281,0.001069833,0.002758766,0.003833469,0.001316678],"domain_scores_gemma":[0.8693148,0.09549685,0.005289643,0.02319024,0.004893978,0.001814498],"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":[0.003551153,0.0009129522,0.09136116,0.001573455,0.0004475589,0.0005747718,0.004933472,0.08854005,0.02626413,0.1392796,0.01572093,0.6268407],"study_design_scores_gemma":[0.0003623245,0.002302401,0.0212833,0.0006628956,0.0003557081,0.001091074,0.002673422,0.508646,0.0616624,0.3789699,0.02179904,0.0001915804],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6648354,0.011731,0.2582882,0.02964913,0.0005914248,0.0002532619,0.0007225861,0.006957401,0.02697161],"genre_scores_gemma":[0.9605796,0.0006585776,0.03522711,0.001039512,0.00009850277,0.00006416914,0.0003740846,0.0004654366,0.001493046],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.02788303,"threshold_uncertainty_score":0.1474613,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02335765223738391,"score_gpt":0.2864959467195923,"score_spread":0.2631382944822084,"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."}}