{"id":"W4221159290","doi":"10.48550/arxiv.2203.13381","title":"Probing Representation Forgetting in Supervised and Unsupervised Continual Learning","year":2022,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Domain Adaptation and Few-Shot Learning","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada; Compute Canada","keywords":"Forgetting; Representation (politics); Computer science; Artificial intelligence; Task (project management); Classifier (UML); Machine learning; Cognitive psychology; Psychology; Engineering","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0007168875,0.0002526494,0.0003218592,0.0004677535,0.0003707204,0.0002382896,0.0008022496,0.0001484148,0.00007536599],"category_scores_gemma":[0.0001642314,0.0003370275,0.0001051595,0.0008179519,0.00006296053,0.0006059069,0.002097766,0.00116145,0.000008781964],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001914684,"about_ca_system_score_gemma":0.0001325518,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002222357,"about_ca_topic_score_gemma":0.00003861393,"domain_scores_codex":[0.9975012,0.0005607142,0.0002939772,0.001132777,0.0001426326,0.000368719],"domain_scores_gemma":[0.9988061,0.0002533146,0.0002530098,0.000491385,0.00007858867,0.0001175584],"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":[0.00005906282,0.00007438528,0.09426131,0.0001216584,0.00005253076,0.0004608524,0.01090136,0.823834,0.0004357811,0.06010741,0.00002804878,0.009663582],"study_design_scores_gemma":[0.001164634,0.00004770741,0.009317992,0.00008093002,0.00001706187,0.000005131023,0.003862791,0.9801866,0.00003885064,0.004161682,0.0007036557,0.0004129695],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7757173,0.00007696803,0.2199641,0.0001453437,0.0002495882,0.000397643,0.000001668234,0.0002728695,0.00317459],"genre_scores_gemma":[0.9948866,0.00008619824,0.003340274,0.00008705168,0.00002993959,0.000004475181,0.00003329165,0.000022301,0.001509855],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2191694,"threshold_uncertainty_score":0.9999081,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07666561706497431,"score_gpt":0.208297278180521,"score_spread":0.1316316611155467,"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."}}