{"id":"W3014594587","doi":"10.1007/978-3-030-58586-0_27","title":"ProxyNCA++: Revisiting and Revitalizing Proxy Neighborhood Component Analysis","year":2020,"lang":"en","type":"preprint","venue":"Lecture notes in computer science","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":19,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Guelph; Vector Institute","funders":"","keywords":"Pooling; Proxy (statistics); Scaling; Metric (unit); Computer science; Similarity (geometry); Recall; Precision and recall; Task (project management); Component (thermodynamics); Artificial intelligence; Algorithm; Machine learning; Pattern recognition (psychology); Mathematics; Image (mathematics); Psychology","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.001461721,0.0005392421,0.0008590865,0.0009770032,0.0004118241,0.001611413,0.003786909,0.0002075811,0.000003209423],"category_scores_gemma":[0.0005197235,0.0005139917,0.0001990702,0.005158195,0.0003486699,0.0004291403,0.007616599,0.001599091,0.00001011343],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002252266,"about_ca_system_score_gemma":0.0003720299,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000332118,"about_ca_topic_score_gemma":0.00001453412,"domain_scores_codex":[0.9946683,0.0002546276,0.0007344768,0.002737457,0.0009102348,0.0006949151],"domain_scores_gemma":[0.9964653,0.0005264598,0.0005373398,0.001930125,0.0002298281,0.0003109021],"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":[0.000003513213,0.00006188924,0.05855365,0.0003435489,0.0001046983,0.0000360776,0.002954626,0.400226,0.001005499,0.003650443,0.000001744645,0.5330583],"study_design_scores_gemma":[0.00013629,0.00003666506,0.04576549,0.0002581152,0.00006204056,0.00001491575,4.863882e-7,0.9351375,0.0003565148,0.01770965,0.00001882104,0.0005035672],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02223119,0.0006233417,0.96363,0.01188134,0.0003073454,0.0008889075,0.000005317157,0.0003970215,0.00003555359],"genre_scores_gemma":[0.5730398,0.00001637005,0.4259906,0.0006754961,0.0002063867,0.0000491244,0.000009155357,0.00001293806,1.364085e-7],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.5508086,"threshold_uncertainty_score":0.9997312,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01792349795523812,"score_gpt":0.2912148204054341,"score_spread":0.2732913224501959,"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."}}