{"id":"W3154476213","doi":"10.18653/v1/2021.cmcl-1.9","title":"TorontoCL at CMCL 2021 Shared Task: RoBERTa with Multi-Stage Fine-Tuning for Eye-Tracking Prediction","year":2021,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":13,"is_retracted":false,"has_abstract":true,"ca_institutions":"Vector Institute; University of Toronto","funders":"Canadian Institute for Advanced Research","keywords":"Computer science; Transformer; Eye tracking; Task (project management); Ranking (information retrieval); Artificial intelligence; Comprehension; Language model; Natural language processing; Tracking (education); Reading comprehension; Task analysis; Machine learning; Reading (process); Programming language; Engineering","routes":{"ca_aff":true,"ca_fund":true,"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":[0.01295763,0.007887378,0.004517665,0.00252898,0.003018051,0.005896039,0.006948558,0.008318134,0.02860794],"category_scores_gemma":[0.03236293,0.001884925,0.003638897,0.002122657,0.00119629,0.004738988,0.007163145,0.006859235,0.04739635],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003647144,"about_ca_system_score_gemma":0.005655313,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.04522753,"about_ca_topic_score_gemma":0.0747695,"domain_scores_codex":[0.9908697,0.003295462,0.000369437,0.003111542,0.00145378,0.000900006],"domain_scores_gemma":[0.9770865,0.005952395,0.0004535498,0.006025375,0.007173074,0.003309173],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.001123156,0.0007892609,0.001815633,0.0005617745,0.0003199805,0.0003557026,0.0002541881,0.006978817,0.0061921,0.0007154024,0.8904746,0.0904193],"study_design_scores_gemma":[0.004684852,0.002378761,0.01625857,0.000490741,0.0006664845,0.001280598,0.00101553,0.5083396,0.03727087,0.01440655,0.4124988,0.0007086272],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"empirical","genre_scores_codex":[0.1260016,0.008958894,0.1273805,0.01685093,0.01722845,0.004055316,0.3651957,0.2989841,0.03534449],"genre_scores_gemma":[0.1684682,0.000770972,0.1427855,0.003356436,0.00225391,0.004009821,0.5965149,0.01682878,0.06501151],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.04522753,"threshold_uncertainty_score":0.09570307,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04593062006633086,"score_gpt":0.2917247730476185,"score_spread":0.2457941529812876,"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."}}