{"id":"W4416091687","doi":"10.48550/arxiv.2505.05335","title":"FLAM: Frame-Wise Language-Audio Modeling","year":2025,"lang":"en","type":"preprint","venue":"ArXiv.org","topic":"Music and Audio Processing","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Event (particle physics); Metadata; JSON; Sample (material); Event data; File format","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0003354017,0.0003822446,0.0004474972,0.0001901288,0.0002126857,0.0003216112,0.002064768,0.0003627748,0.00004458701],"category_scores_gemma":[0.0001139288,0.0003736118,0.000205021,0.00032807,0.0000442029,0.00029191,0.003107176,0.001043527,0.0001846159],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00008576371,"about_ca_system_score_gemma":0.0004741867,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001501953,"about_ca_topic_score_gemma":0.00001262686,"domain_scores_codex":[0.9976182,0.00006814388,0.0004309824,0.001054895,0.0003478657,0.0004799172],"domain_scores_gemma":[0.9979826,0.00006249919,0.0002077667,0.001478099,0.0001435537,0.0001254504],"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":[0.00005800453,0.0007346809,0.1009124,0.005774911,0.000947574,0.0009754088,0.04799236,0.2683685,0.006338204,0.01493789,0.03391092,0.5190492],"study_design_scores_gemma":[0.000483459,0.00002330821,0.00173229,0.001787302,0.00007930445,0.00001079055,0.0001800711,0.9765176,0.004544371,0.008764543,0.004597791,0.001279178],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3455666,0.001922647,0.640412,0.001745678,0.001389871,0.0001630846,0.00000563919,0.0004547316,0.008339717],"genre_scores_gemma":[0.959121,0.0001142582,0.03221749,0.003615054,0.0005721459,0.00004604015,0.00001944927,0.00002366269,0.004270861],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.7081491,"threshold_uncertainty_score":0.9998716,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04744947382599204,"score_gpt":0.2921131689902425,"score_spread":0.2446636951642504,"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."}}