{"id":"W4417442172","doi":"10.21468/scipostphyscore.8.4.095","title":"Tensorization of neural networks for improved privacy and interpretability","year":2025,"lang":"en","type":"preprint","venue":"SciPost Physics Core","topic":"Generative Adversarial Networks and Image Synthesis","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Agencia Estatal de Investigación; Government of Canada; Ministry of Colleges and Universities; Ministerio de Ciencia e Innovación; Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung; European Commission; Ministerio de Asuntos Económicos y Transformación Digital, Gobierno de España; NextGenerationEU; National Science Foundation; Universidad Complutense de Madrid; Comunidad de Madrid; Institut Périmètre de physique théorique; Innovation, Science and Economic Development Canada","keywords":"Interpretability; Artificial neural network; Initialization; Domain (mathematical analysis); ENCODE; Set (abstract data type); Function (biology); Interpolation (computer graphics); Feature (linguistics)","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":[0.001853322,0.001244193,0.0008604029,0.0007284071,0.0005039746,0.001492228,0.001252585,0.001044361,0.003465457],"category_scores_gemma":[0.008117015,0.0005788421,0.0009368651,0.0008214191,0.001949492,0.003276911,0.003055001,0.003236743,0.0008820557],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001342193,"about_ca_system_score_gemma":0.001268378,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001857682,"about_ca_topic_score_gemma":0.002135182,"domain_scores_codex":[0.9990312,0.0003802807,0.00005140188,0.0001985231,0.0002374085,0.000101186],"domain_scores_gemma":[0.9973803,0.001112245,0.0002574715,0.0009372688,0.000219039,0.0000935832],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0002117033,0.00005022072,0.0006803646,0.0001109809,0.00005339237,0.0001289247,0.000185785,0.5454196,0.01375499,0.3051641,0.003012051,0.131228],"study_design_scores_gemma":[0.000005492914,0.00001811952,0.00004443055,0.000006156326,0.00000453797,0.000023616,0.000007523225,0.9312054,0.003379634,0.06441421,0.0008838324,0.000007039077],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.005542018,0.00005486242,0.9930505,0.0001372189,0.00001315975,0.00001358099,0.00006031023,0.0003766899,0.0007518181],"genre_scores_gemma":[0.4825709,0.0004363386,0.5096004,0.0002931482,0.00009698202,0.0001928438,0.0005071023,0.0006299791,0.005672249],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.003465457,"threshold_uncertainty_score":0.01159304,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02478107824929196,"score_gpt":0.2733115271267021,"score_spread":0.2485304488774102,"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."}}