{"id":"W6966597627","doi":"10.48448/m47m-ta93","title":"A Systematic Survey and Critical Review on Evaluating Large Language Models: Challenges, Limitations, and Recommendations","year":2024,"lang":"en","type":"other","venue":"Underline Science Inc.","topic":"","field":"","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"York University","funders":"","keywords":"Process (computing); Systematic review; Evaluation methods; Reliability (semiconductor); Comparability","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":["metaresearch","metaepi_narrow","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0126907,0.0004013291,0.0007529957,0.0009261789,0.0002391895,0.0002927617,0.0003798994,0.0001458012,0.0001634783],"category_scores_gemma":[0.0119841,0.0003314994,0.00004237963,0.0009566501,0.0005546177,0.0002816327,0.0002811987,0.0004349147,0.001193577],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001656254,"about_ca_system_score_gemma":0.0003118294,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001044589,"about_ca_topic_score_gemma":0.002440506,"domain_scores_codex":[0.9957467,0.00100307,0.0006420813,0.001077347,0.001020437,0.0005103229],"domain_scores_gemma":[0.9967988,0.001715735,0.0002621108,0.0007129096,0.0002665808,0.0002438671],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"systematic_review","study_design_scores_codex":[0.00001073194,0.0005778489,0.00000608543,0.3454646,0.0002704912,0.00004688056,0.004165618,0.00001067926,0.00006101027,0.2647046,0.3675757,0.0171058],"study_design_scores_gemma":[0.0009386324,0.0006571272,0.00005044737,0.6484549,0.001489513,0.0002065192,0.007041179,0.3158585,0.000002180467,0.01144111,0.01169615,0.002163716],"study_design_candidate":"systematic_review","study_design_consensus":null,"genre_codex":"review","genre_gemma":"review","genre_scores_codex":[0.0000209818,0.8565112,0.0006515994,0.005158274,0.0005091832,0.002700338,0.001757935,0.0006334658,0.1320571],"genre_scores_gemma":[0.06668884,0.6289054,0.08062037,0.006406014,0.00112952,0.002281232,0.002507925,0.005834048,0.2056267],"genre_candidate":"review","genre_consensus":"review","teacher_disagreement_score":0.3558795,"threshold_uncertainty_score":0.9999137,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2749499417627526,"score_gpt":0.4398776242197662,"score_spread":0.1649276824570136,"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."}}