{"id":"W4385436799","doi":"10.48550/arxiv.2307.15508","title":"The Road to Quality is Paved with Good Revisions: A Detailed Evaluation Methodology for Revision Policies in Incremental Sequence Labelling","year":2023,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Speech and dialogue systems","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Atomic Energy of Canada Limited","keywords":"Computer science; Prefix; Sequence (biology); Labelling; Encoder; Quality (philosophy); Transformer; Engineering","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.03696633,0.001468954,0.001402682,0.003507935,0.00115106,0.004875186,0.002743836,0.002691092,0.001850223],"category_scores_gemma":[0.1825025,0.0009141341,0.0009671108,0.002369198,0.003115596,0.00743335,0.003710253,0.003302887,0.0006586924],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003898358,"about_ca_system_score_gemma":0.004215342,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006564101,"about_ca_topic_score_gemma":0.007078425,"domain_scores_codex":[0.9565759,0.02270363,0.004363593,0.00424161,0.01084014,0.001275119],"domain_scores_gemma":[0.8043579,0.1196904,0.01446017,0.0306297,0.02661049,0.004251375],"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":[0.002064555,0.0008168959,0.03500763,0.001289529,0.0004837557,0.0002383675,0.003692977,0.2795123,0.02609744,0.06097334,0.005224807,0.5845985],"study_design_scores_gemma":[0.0001057027,0.00127954,0.005973633,0.000165332,0.0001403243,0.0002697989,0.0003943419,0.9134632,0.03027099,0.04311432,0.004668292,0.0001545901],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.06971976,0.0006845841,0.9215406,0.0005982638,0.00009689932,0.0005092267,0.0004764462,0.003799599,0.002574626],"genre_scores_gemma":[0.5629856,0.0002531366,0.4337156,0.0001819533,0.00005359002,0.0003258286,0.0007421185,0.000732915,0.001009178],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.03696633,"threshold_uncertainty_score":0.195499,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.4732101525060852,"score_gpt":0.3479648132122058,"score_spread":0.1252453392938793,"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."}}