{"id":"W2969211464","doi":"10.1109/iccv.2019.00040","title":"Overcoming Catastrophic Forgetting With Unlabeled Data in the Wild","year":2019,"lang":"en","type":"preprint","venue":"","topic":"Domain Adaptation and Few-Shot Learning","field":"Computer Science","cited_by":23,"is_retracted":false,"has_abstract":true,"ca_institutions":"Kootenay Association for Science & Technology","funders":"Kwanjeong Educational Foundation; National Science Foundation","keywords":"Forgetting; Leverage (statistics); Overfitting; Computer science; Artificial intelligence; Machine learning; Code (set theory); Labeled data; Distillation; Artificial neural network; Task (project management); Generalization; Mathematics; Engineering","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":[],"consensus_categories":[],"category_scores_codex":[0.001463452,0.0002337527,0.0002526256,0.0001397953,0.0001061473,0.0006089347,0.004820599,0.0001046507,0.00001558988],"category_scores_gemma":[0.0001061987,0.0001500797,0.00003328372,0.0003210719,0.00003293431,0.0005026332,0.003379801,0.001014011,0.00005406289],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004479718,"about_ca_system_score_gemma":0.0002503829,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002871751,"about_ca_topic_score_gemma":0.0001571097,"domain_scores_codex":[0.9977823,0.000216107,0.0003047684,0.0008387533,0.0005100735,0.0003480216],"domain_scores_gemma":[0.9963021,0.0003118997,0.0002291408,0.003071003,0.00004238962,0.00004344909],"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":[0.0001520023,0.0006569622,0.1061058,0.001101531,0.0004648429,0.0005828527,0.03536623,0.2808364,0.0003421688,0.277207,0.01554248,0.2816417],"study_design_scores_gemma":[0.0007711436,0.00005070862,0.008467419,0.0003158384,0.00001581862,0.00003530672,0.0007057055,0.9737865,0.00001296574,0.001062608,0.01428389,0.0004920581],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01646107,0.0001259209,0.9698252,0.00301289,0.0003372777,0.0005418973,0.000008652467,0.0001420549,0.009545008],"genre_scores_gemma":[0.8519005,0.00001180507,0.145883,0.001486507,0.00007469719,0.00001745457,0.0001280738,0.000017823,0.0004801919],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8354394,"threshold_uncertainty_score":0.8957952,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04958758441825139,"score_gpt":0.281441131426253,"score_spread":0.2318535470080016,"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."}}