{"id":"W4408352598","doi":"10.1109/icassp49660.2025.10889047","title":"Learning from Ambiguous Data with Hard Labels","year":2025,"lang":"en","type":"article","venue":"","topic":"Machine Learning and Data Classification","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"","keywords":"Computer science; Artificial intelligence","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":[0.0002039371,0.00007262664,0.00008334124,0.00005396747,0.0001243241,0.0002449412,0.001387843,0.00002773685,0.00004112607],"category_scores_gemma":[0.0001072727,0.00005436309,0.000007490827,0.0002936615,0.00001837284,0.0004642623,0.0005634149,0.0001858089,0.0001327474],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001017993,"about_ca_system_score_gemma":0.00006192747,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001238994,"about_ca_topic_score_gemma":0.00007251799,"domain_scores_codex":[0.9991131,0.0000726237,0.0000990848,0.0004589435,0.0001348182,0.0001214327],"domain_scores_gemma":[0.998432,0.00007153391,0.00004240086,0.001390281,0.00003134275,0.00003239857],"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":[0.00001482173,0.00007052426,0.04039936,0.0000128414,0.00005945967,0.00001016774,0.0002429635,0.0005187105,0.001253872,0.08475874,0.01935981,0.8532987],"study_design_scores_gemma":[0.000457455,0.0000688134,0.1623619,0.00004410617,0.00001495229,0.000002118028,0.0000661782,0.5698636,0.0003714287,0.001004194,0.2655481,0.0001971303],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.009469145,0.00006149911,0.963396,0.004496992,0.00009966688,0.00004845993,0.000004898076,0.0003969476,0.02202638],"genre_scores_gemma":[0.8028298,0.00001895431,0.1769267,0.0009492372,0.00003973559,0.000005278216,0.0004034725,0.000006101046,0.01882073],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8531016,"threshold_uncertainty_score":0.257898,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03455206419651542,"score_gpt":0.2792409356508822,"score_spread":0.2446888714543668,"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."}}