{"id":"W3169315049","doi":"10.1007/s10994-021-06080-w","title":"On the benefits of representation regularization in invariance based domain generalization","year":2022,"lang":"en","type":"article","venue":"Machine Learning","topic":"Domain Adaptation and Few-Shot Learning","field":"Computer Science","cited_by":20,"is_retracted":false,"has_abstract":true,"ca_institutions":"Canadian Institute for Advanced Research; Mila - Quebec Artificial Intelligence Institute; Vector Institute; Western University; Université Laval","funders":"Canadian Network for Research and Innovation in Machining Technology, Natural Sciences and Engineering Research Council of Canada; Natural Sciences and Engineering Research Council of Canada; Mitacs; Canadian Institute for Advanced Research","keywords":"Regularization (linguistics); Invariant (physics); Computer science; Generalization error; Artificial intelligence; Representation (politics); Generalization; Machine learning; Feature learning; Robustness (evolution); Algorithm; Mathematics; Unsupervised learning","routes":{"ca_aff":true,"ca_fund":true,"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":[0.001056149,0.00008364351,0.000101181,0.0002024455,0.000349006,0.0000571128,0.0003733576,0.00002152049,0.0001388951],"category_scores_gemma":[0.0003238167,0.00007581317,0.00003388252,0.001100514,0.00001701498,0.0001598003,0.0001341104,0.0002777131,0.000003929924],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006198621,"about_ca_system_score_gemma":0.00003599824,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006140406,"about_ca_topic_score_gemma":0.00001497051,"domain_scores_codex":[0.9980544,0.0008965424,0.0002496894,0.0002560438,0.0004107815,0.0001325153],"domain_scores_gemma":[0.9991592,0.0002707758,0.0002374675,0.0002742243,0.00003777714,0.00002054023],"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":[0.00001256829,0.0000236936,0.00471676,0.000002690027,0.000001849165,0.000001249844,0.0009762855,0.6889799,0.0005977808,0.3005989,0.00002479788,0.00406345],"study_design_scores_gemma":[0.0004554857,0.00008370064,0.01054604,0.00001315497,0.000001379657,0.000001853198,0.0001174061,0.979813,0.0003992792,0.006888104,0.001593934,0.00008659095],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1252363,0.00007076799,0.8678458,0.003597886,0.0001394008,0.000232698,0.00000156947,0.0000865542,0.002789021],"genre_scores_gemma":[0.9896374,0.000003103518,0.0088507,0.0008406352,0.00001472963,0.00004173554,0.00005185059,0.00001133828,0.0005485404],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8644011,"threshold_uncertainty_score":0.3091572,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02131954762480759,"score_gpt":0.2423029124869489,"score_spread":0.2209833648621413,"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."}}