{"id":"W4376615930","doi":"10.1002/smr.2571","title":"CodeBERT‐Attack: Adversarial attack against source code deep learning models via pre‐trained model","year":2023,"lang":"en","type":"article","venue":"Journal of Software Evolution and Process","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":17,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"JST-Mirai Program; Japan Society for the Promotion of Science; Natural Sciences and Engineering Research Council of Canada; National Natural Science Foundation of China; Canadian Institute for Advanced Research","keywords":"Computer science; Source code; Adversarial system; Code (set theory); Artificial intelligence; Machine learning; Masking (illustration); Vulnerability (computing); Deep learning; Inference; Programming language; Computer security","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001269575,0.001233951,0.0005744321,0.0004472035,0.0003566431,0.0005192652,0.001099745,0.001190841,0.001424758],"category_scores_gemma":[0.006136385,0.0003554132,0.0007378841,0.0002084569,0.001198441,0.001469472,0.001857733,0.002735857,0.0004277574],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008251989,"about_ca_system_score_gemma":0.0008848555,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002823728,"about_ca_topic_score_gemma":0.002436701,"domain_scores_codex":[0.999083,0.0002983527,0.00004413863,0.0001733172,0.0002694009,0.0001316934],"domain_scores_gemma":[0.9971555,0.001584665,0.0002514515,0.0006152563,0.0002991405,0.00009394604],"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":[0.000282242,0.00007218385,0.002156343,0.00007281038,0.00008282899,0.0002233508,0.00008801945,0.9066708,0.01442086,0.00828963,0.002681918,0.06495898],"study_design_scores_gemma":[0.000002919361,0.00002327401,0.00006360069,0.000003768973,0.000004237324,0.00001760868,0.000002992684,0.9952141,0.003018155,0.001487128,0.0001591569,0.000003025594],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1538048,0.000553445,0.8354025,0.001132406,0.0001483901,0.0001037981,0.0001249643,0.005327948,0.003401619],"genre_scores_gemma":[0.950242,0.000123567,0.04681069,0.0004076789,0.0000292074,0.00005878315,0.0001406731,0.0001619894,0.00202539],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.002823728,"threshold_uncertainty_score":0.006714284,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02888597047409828,"score_gpt":0.289055668582603,"score_spread":0.2601696981085047,"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."}}