{"id":"W4400525527","doi":"10.1145/3626772.3657675","title":"Towards Robust QA Evaluation via Open LLMs","year":2024,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":14,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"Universitas Brawijaya","keywords":"Computer science","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":[0.0008115367,0.00005782618,0.00005991797,0.00004759205,0.00004027433,0.0006991193,0.0009410783,0.00002857241,0.0004041143],"category_scores_gemma":[0.00001858396,0.00004639204,0.00002399996,0.0002067699,0.000005523419,0.0007543758,0.0005423117,0.00006116766,0.0002837319],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006034725,"about_ca_system_score_gemma":0.0001584829,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002619164,"about_ca_topic_score_gemma":0.00003071947,"domain_scores_codex":[0.9991096,0.0000383611,0.0001178384,0.000313571,0.0003012701,0.0001193651],"domain_scores_gemma":[0.9994857,0.00001859723,0.00001043444,0.0003898031,0.00005659013,0.00003883114],"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":[3.471544e-7,0.000007866299,0.00001056913,0.000007647873,0.000007314213,0.00000437132,0.0003563892,0.007027805,0.0001566816,0.1479867,0.002751876,0.8416825],"study_design_scores_gemma":[0.00006367886,0.00001137189,0.0001136339,0.0000134541,0.000004013241,0.000006302002,0.000007128157,0.9721294,0.0004567189,0.02247013,0.004654921,0.00006924237],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.001095033,0.0001577759,0.9079497,0.002893195,0.0007603765,0.0001987572,1.648167e-7,0.0001892325,0.08675577],"genre_scores_gemma":[0.708904,0.000003474719,0.2872782,0.0004511078,0.0001121347,0.00003092817,0.000001557422,0.000005848663,0.003212798],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9651016,"threshold_uncertainty_score":0.6741623,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1242644351317595,"score_gpt":0.3507526136183011,"score_spread":0.2264881784865415,"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."}}