{"id":"W4412534222","doi":"10.1088/2632-2153/adf278","title":"Feature learning and generalization in deep networks with orthogonal weights","year":2025,"lang":"en","type":"article","venue":"Machine Learning Science and Technology","topic":"Neural Networks and Applications","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Vector Institute; University of Toronto","funders":"High Energy Physics","keywords":"Generalization; Feature (linguistics); Artificial intelligence; Pattern recognition (psychology); Feature learning; Computer science; Deep learning; Mathematics; Philosophy","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":[0.0008829096,0.0003816636,0.0003016304,0.0003210023,0.0001987159,0.0004507874,0.0007493729,0.0004476481,0.0009208443],"category_scores_gemma":[0.004325534,0.0003247713,0.0003079021,0.0002785566,0.0008059688,0.001409866,0.001024604,0.0009631715,0.0001207954],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007983103,"about_ca_system_score_gemma":0.0004882021,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002457575,"about_ca_topic_score_gemma":0.001809518,"domain_scores_codex":[0.9998226,0.00003774242,0.000009480343,0.00004878455,0.00004055083,0.00004085449],"domain_scores_gemma":[0.9990283,0.0004275317,0.0001584203,0.0001941354,0.0001331938,0.00005853346],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.000171762,0.00008884556,0.002940273,0.00005552438,0.00004455369,0.00009281979,0.0001271715,0.860735,0.03818618,0.03836403,0.0007835428,0.05841033],"study_design_scores_gemma":[0.000003811466,0.00002397648,0.0003285341,0.000003550201,0.000002565959,0.000008843163,0.000004332359,0.9886969,0.003775185,0.007041452,0.0001075055,0.000003241468],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4831277,0.0001511015,0.5129089,0.0002611096,0.00002577764,0.00002459551,0.00008583409,0.001029958,0.00238502],"genre_scores_gemma":[0.9693686,0.0000367741,0.029888,0.00003741374,0.0000051889,0.00001718699,0.00006266143,0.00004134572,0.0005427404],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.002457575,"threshold_uncertainty_score":0.005792201,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.002308074299525024,"score_gpt":0.2194735595103882,"score_spread":0.2171654852108632,"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."}}