{"id":"W7111079672","doi":"10.48550/arxiv.2512.05998","title":"Going All-In on LLM Accuracy: Fake Prediction Markets, Real Confidence Signals","year":2025,"lang":"","type":"preprint","venue":"arXiv (Cornell University)","topic":"Explainable Artificial Intelligence (XAI)","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"","keywords":"Incentive; Framing (construction); Baseline (sea); Binary number; Scoring rule; Framing effect; Confidence interval; Task (project management)","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":["metaepi_narrow","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.002409036,0.001119513,0.001085286,0.001546074,0.0006202819,0.0006616164,0.00518695,0.001017103,0.0003972059],"category_scores_gemma":[0.0008714496,0.001490213,0.000552651,0.003154651,0.0004842499,0.002078276,0.003516137,0.002256385,0.0008938472],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001863177,"about_ca_system_score_gemma":0.001498737,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00242997,"about_ca_topic_score_gemma":0.0009528711,"domain_scores_codex":[0.990922,0.001376916,0.001326998,0.00422517,0.0005066114,0.001642256],"domain_scores_gemma":[0.991775,0.002792282,0.00103738,0.003189236,0.0006737008,0.000532445],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0006179247,0.0007368929,0.004915706,0.0003082282,0.000223068,0.001886097,0.001715873,0.5925948,0.0004876143,0.3893751,0.0006793624,0.006459399],"study_design_scores_gemma":[0.000555825,0.0004176354,0.004273589,0.002281464,0.0001575774,0.00000957529,0.0007869534,0.931662,0.005719052,0.05002861,0.002874966,0.001232718],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3529845,0.0001577963,0.5830487,0.0006860236,0.003941125,0.002398181,0.000149086,0.0006094105,0.0560252],"genre_scores_gemma":[0.9824389,0.005154638,0.0004865668,0.0004658438,0.0001892159,0.000009624501,0.00002980345,0.00003730658,0.01118808],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.6294544,"threshold_uncertainty_score":0.9998841,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1454807877301481,"score_gpt":0.2492995981184722,"score_spread":0.1038188103883241,"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."}}