{"id":"W4400526764","doi":"10.1145/3626772.3657824","title":"MTMS: Multi-teacher Multi-stage Knowledge Distillation for Reasoning-Based Machine Reading Comprehension","year":2024,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":8,"is_retracted":false,"has_abstract":true,"ca_institutions":"York University","funders":"Universitas Brawijaya","keywords":"Computer science; Distillation; Comprehension; Artificial intelligence; Reading (process); Reading comprehension; Natural language processing; Stage (stratigraphy); Programming language; Linguistics; Chemistry","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":[0.0009107036,0.001470056,0.001003823,0.0005701284,0.0006214743,0.001027699,0.002790299,0.001398697,0.006943345],"category_scores_gemma":[0.004424005,0.0005902481,0.0009933094,0.0005878872,0.0008744356,0.002987402,0.003170373,0.003113791,0.002291785],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007977429,"about_ca_system_score_gemma":0.002050983,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005178288,"about_ca_topic_score_gemma":0.0113254,"domain_scores_codex":[0.9993746,0.0001910765,0.00003854519,0.0001979277,0.0001329551,0.00006487115],"domain_scores_gemma":[0.9989787,0.0005638722,0.000067407,0.0001859685,0.0001195305,0.0000844942],"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":[0.0006409019,0.0004268346,0.001503574,0.0006622813,0.0001671072,0.0003231277,0.0005059778,0.2385265,0.0231091,0.02222226,0.01910749,0.6928048],"study_design_scores_gemma":[0.00005708901,0.00009256798,0.0001472292,0.00001929611,0.00001887583,0.00004289639,0.00003315519,0.9746882,0.00728645,0.01341193,0.004181853,0.00002049012],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01853815,0.001148505,0.9583548,0.0006748078,0.0002405501,0.0001558401,0.0005306878,0.01734971,0.00300702],"genre_scores_gemma":[0.4384035,0.0005419055,0.5485939,0.0006866613,0.0001949787,0.0004460421,0.002302237,0.001081955,0.007748766],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.006943345,"threshold_uncertainty_score":0.02322781,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08466576282071729,"score_gpt":0.3485421448935959,"score_spread":0.2638763820728786,"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."}}