{"id":"W4402671693","doi":"10.18653/v1/2024.acl-long.680","title":"To Distill or Not to Distill? On the Robustness of Robust Knowledge Distillation","year":2024,"lang":"en","type":"article","venue":"","topic":"Bayesian Modeling and Causal Inference","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Alliance de recherche numérique du Canada; Social Sciences and Humanities Research Council of Canada; Natural Sciences and Engineering Research Council of Canada; Canada Research Chairs","keywords":"Robustness (evolution); Computer science; Distillation; Artificial intelligence; Machine learning; Chromatography; Biology; Chemistry","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":[],"consensus_categories":[],"category_scores_codex":[0.0003576817,0.0001802805,0.0001780229,0.0001231395,0.00009958396,0.0002759226,0.0008398417,0.0000545936,0.00008098847],"category_scores_gemma":[0.0001893354,0.0000957151,0.00006653487,0.0009593009,0.00003535532,0.0001502998,0.0002877568,0.0001357639,0.0002965862],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005691611,"about_ca_system_score_gemma":0.0001066012,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003362101,"about_ca_topic_score_gemma":0.00008670351,"domain_scores_codex":[0.9986142,0.00005983024,0.0002849009,0.0005015698,0.0002814823,0.000258026],"domain_scores_gemma":[0.9985695,0.0004323414,0.0000291677,0.0006687976,0.0001364307,0.0001637343],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00007226274,0.0001089808,0.00002039882,0.00009436671,0.00002435148,0.000009205613,0.001910315,0.1252837,0.0003956352,0.7083004,0.019916,0.1438644],"study_design_scores_gemma":[0.00004948047,0.0002091663,0.001110716,0.0002404641,0.000008824193,0.000004569658,0.0000424305,0.9899217,0.001460145,0.0006356732,0.006078481,0.000238282],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01624257,0.0000234462,0.9706197,0.006657079,0.0007487535,0.0002475687,0.00001539159,0.0002372748,0.005208166],"genre_scores_gemma":[0.9803825,0.000003233441,0.01446475,0.0004015875,0.0001350573,0.00002834209,0.000002067993,0.00001400707,0.004568482],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9641399,"threshold_uncertainty_score":0.3903148,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08253438536572152,"score_gpt":0.3196380303735552,"score_spread":0.2371036450078337,"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."}}