{"id":"W7134163353","doi":"10.1109/bigdata66926.2025.11400965","title":"Enhancing Interpretability in Self-Training with Tsetlin Machines for Mitigating Noisy Pseudo-Labels","year":2025,"lang":"","type":"article","venue":"","topic":"Machine Learning and Data Classification","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Toronto","funders":"","keywords":"Interpretability; Noisy data; Feature (linguistics); Noise (video); Pattern recognition (psychology)","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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.003227469,0.0004964468,0.0006524607,0.000433144,0.0004687864,0.0006986479,0.00122171,0.0002046489,0.00004518554],"category_scores_gemma":[0.001431598,0.0004367304,0.000131227,0.001497361,0.0001249492,0.0009787189,0.0004768403,0.000747725,0.00001301914],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002346509,"about_ca_system_score_gemma":0.0007635626,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0005473508,"about_ca_topic_score_gemma":0.002427182,"domain_scores_codex":[0.9955492,0.0004682396,0.001223642,0.00158236,0.0003411036,0.0008354431],"domain_scores_gemma":[0.9962745,0.001652788,0.0004008104,0.001250164,0.0002734635,0.0001483007],"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":[0.0002562853,0.0007481748,0.1276033,0.002205804,0.0002164397,0.000009374025,0.03115854,0.001396262,0.01125333,0.1135732,0.0001660656,0.7114133],"study_design_scores_gemma":[0.001434714,0.0002741189,0.01552409,0.001221138,0.00005230825,0.000008317062,0.0009545542,0.9732911,0.002891713,0.003093212,0.000742199,0.0005125193],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1463626,0.0001940422,0.8405792,0.004243945,0.0005483213,0.0009925548,0.00001723478,0.0003482732,0.006713861],"genre_scores_gemma":[0.7107927,0.000009158415,0.2877694,0.0005571094,0.00006809081,0.0001300887,0.00002590442,0.00001961104,0.0006279438],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9718949,"threshold_uncertainty_score":0.9998084,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0142521620139618,"score_gpt":0.2914397057257438,"score_spread":0.277187543711782,"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."}}