{"id":"W6979343015","doi":"","title":"Model Compression vs. Adversarial Robustness: An Empirical Study on Language Models for Code","year":2025,"lang":"en","type":"article","venue":"ArXiv.org","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Adversarial system; Software; Source lines of code; Robustness (evolution); Source code; Software deployment; Uncompressed video; Empirical research; Language model","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00946792,0.001121407,0.0007235269,0.001392829,0.0007127717,0.001294875,0.001285056,0.001243348,0.001503856],"category_scores_gemma":[0.08976205,0.0003681375,0.0007974142,0.001133553,0.002564162,0.005065795,0.001853601,0.002972722,0.0004834019],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001814179,"about_ca_system_score_gemma":0.00134339,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004264922,"about_ca_topic_score_gemma":0.003490166,"domain_scores_codex":[0.9946307,0.002501148,0.0003088973,0.0007680838,0.001472179,0.0003189953],"domain_scores_gemma":[0.9003124,0.08289227,0.00450746,0.008892735,0.002678257,0.0007167524],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"observational","study_design_scores_codex":[0.001281118,0.0005430522,0.03841238,0.0007613477,0.0003283281,0.000432961,0.0006187161,0.8220236,0.004796533,0.01746578,0.009645292,0.1036909],"study_design_scores_gemma":[0.00004479599,0.0003601497,0.005122962,0.00008322583,0.00005819399,0.0003052426,0.0001810991,0.9745513,0.003716787,0.01358484,0.0019533,0.00003815632],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8755379,0.003269581,0.1097765,0.00226867,0.0001597151,0.0003373454,0.001853338,0.001850694,0.004946204],"genre_scores_gemma":[0.9830014,0.0005013614,0.0136143,0.0002022778,0.00004507406,0.0001021505,0.001785631,0.0001848447,0.0005629882],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.00946792,"threshold_uncertainty_score":0.05007172,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07452659093562705,"score_gpt":0.3712633562769646,"score_spread":0.2967367653413375,"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."}}