{"id":"W4404007859","doi":"10.1101/2024.10.31.621071","title":"WhaleLM: Finding Structure and Information in Sperm Whale Vocalizations and Behavior with Machine Learning","year":2024,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Marine animal studies overview","field":"Environmental Science","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"Carleton University","funders":"","keywords":"Sperm whale; Whale; Sperm; Artificial intelligence; Communication; Biology; Computer science; Speech recognition; Fishery; Psychology; Botany","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0005454146,0.0002533532,0.0001897787,0.0003954564,0.0001840534,0.0003619606,0.0002803812,0.0004775542,0.001158366],"category_scores_gemma":[0.003289815,0.0002069897,0.0002934341,0.0001984398,0.0003067393,0.0004710796,0.0003915374,0.0004538344,0.0002568176],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002728023,"about_ca_system_score_gemma":0.0001631388,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005522062,"about_ca_topic_score_gemma":0.006222578,"domain_scores_codex":[0.9998295,0.00006396284,0.000005532416,0.00005979849,0.0000233548,0.00001778299],"domain_scores_gemma":[0.9986542,0.0009514118,0.0001494632,0.0001031054,0.00007482042,0.0000670572],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0006355949,0.0005598658,0.3931386,0.0002104856,0.0004428139,0.0003325442,0.001137912,0.3161975,0.1743994,0.003065561,0.001968612,0.1079113],"study_design_scores_gemma":[0.000004847042,0.000101485,0.09882849,0.00000869057,0.00001487122,0.00004137788,0.0001032448,0.8916373,0.008337392,0.0007553484,0.0001482394,0.00001874024],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9845806,0.00004242186,0.01458881,0.00008590562,0.000005677693,0.000006801578,0.0001141602,0.0001430784,0.0004326174],"genre_scores_gemma":[0.9953532,0.0000109472,0.004162592,0.00001184369,0.000003728011,0.000007423176,0.0001763776,0.00001614726,0.0002577503],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.005522062,"threshold_uncertainty_score":0.01097983,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.008201748112892477,"score_gpt":0.2025444645846058,"score_spread":0.1943427164717133,"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."}}