{"id":"W4389520674","doi":"10.18653/v1/2023.emnlp-main.433","title":"Unified Low-Resource Sequence Labeling by Sample-Aware Dynamic Sparse Finetuning","year":2023,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"Cisco Systems","keywords":"Computer science; Leverage (statistics); Sequence (biology); Context (archaeology); Resource (disambiguation); Sample (material); Generalization; Process (computing); Artificial intelligence; Machine learning; Data mining; Programming language","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":[0.0003817064,0.0001751992,0.0001674125,0.0001394847,0.0001834681,0.0001807877,0.001131785,0.00007639406,0.00004648443],"category_scores_gemma":[0.0001174195,0.0001715732,0.00004939563,0.0008311726,0.00002934583,0.0003181719,0.0005176182,0.0001995645,0.0002257244],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006833775,"about_ca_system_score_gemma":0.00006733003,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000276816,"about_ca_topic_score_gemma":0.00005165818,"domain_scores_codex":[0.99823,0.00005660761,0.0002821286,0.0005776872,0.0003349702,0.0005186392],"domain_scores_gemma":[0.9986619,0.0003034773,0.00006673887,0.0007859449,0.00005257245,0.0001293756],"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":[0.00001884271,0.0001462482,0.003287396,0.0003150599,0.0001255191,0.0004010627,0.008252648,0.3760413,0.08441128,0.1026955,0.03566166,0.3886434],"study_design_scores_gemma":[0.0001663941,0.00001307418,0.00004322707,0.00004483398,0.000002439175,0.000004289109,0.0001364729,0.9931383,0.0007525112,0.001623814,0.003849515,0.0002250871],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.08531328,0.00004289683,0.910703,0.001664885,0.0001978004,0.0001145134,0.00001121451,0.001257146,0.0006952298],"genre_scores_gemma":[0.9170918,0.00001983908,0.07783134,0.0009224874,0.00004145637,0.00001328124,0.00005123116,0.00002452605,0.004004064],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8328717,"threshold_uncertainty_score":0.6996554,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04192559110134546,"score_gpt":0.2735007169432181,"score_spread":0.2315751258418727,"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."}}