{"id":"W4416136027","doi":"10.48550/arxiv.2506.05027","title":"Tuning the Right Foundation Models is What you Need for Partial Label Learning","year":2025,"lang":"en","type":"preprint","venue":"ArXiv.org","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Canadian Institute for Advanced Research","keywords":"Ambiguity; Initialization; Foundation (evidence); Benchmark (surveying); Classifier (UML); Convolutional neural network; Model selection","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":[0.007079498,0.002012754,0.001524114,0.001185264,0.001457141,0.003312396,0.003191874,0.003006391,0.004413584],"category_scores_gemma":[0.02209309,0.0007879058,0.001549494,0.001298544,0.002835804,0.01511099,0.004203607,0.007172881,0.006071779],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001537173,"about_ca_system_score_gemma":0.002206031,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004462334,"about_ca_topic_score_gemma":0.01084767,"domain_scores_codex":[0.9964584,0.001477587,0.0001663215,0.001030902,0.0006176116,0.0002493243],"domain_scores_gemma":[0.9877619,0.003798723,0.0005646868,0.006052546,0.00129514,0.00052706],"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":[0.0006275628,0.0005297248,0.01234598,0.0007931733,0.0003294338,0.0001950429,0.0008363671,0.08022258,0.01471275,0.04910922,0.09299237,0.7473059],"study_design_scores_gemma":[0.0001281581,0.0003656831,0.002050052,0.0004894572,0.000123182,0.0003992186,0.0008552909,0.6809211,0.01211715,0.245612,0.05678435,0.000154319],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.03093853,0.002451486,0.9378113,0.008496846,0.0004428404,0.0001954579,0.001299627,0.01280273,0.005561269],"genre_scores_gemma":[0.2463891,0.001546573,0.733588,0.004518217,0.000440808,0.0004541489,0.005842989,0.002557621,0.004662552],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007079498,"threshold_uncertainty_score":0.03744042,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08464897028036827,"score_gpt":0.3101483991326191,"score_spread":0.2254994288522508,"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."}}