{"id":"W4385485199","doi":"10.1109/compsac57700.2023.00084","title":"Milo: Attacking Deep Pre-trained Model for Programming Languages Tasks with Anti-analysis Code Obfuscation","year":2023,"lang":"en","type":"article","venue":"","topic":"Advanced Malware Detection Techniques","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":false,"ca_institutions":"Queen's University","funders":"","keywords":"Obfuscation; Computer science; Programming language; Code (set theory); Natural language processing; Artificial intelligence; Computer security","routes":{"ca_aff":true,"ca_fund":false,"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":[0.0008227528,0.002499391,0.0009132526,0.0009075989,0.0005862866,0.0009826148,0.001801661,0.002015593,0.006001786],"category_scores_gemma":[0.002829721,0.000767917,0.001338036,0.0003835824,0.0005122874,0.001982168,0.001874839,0.003054434,0.00367271],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007654123,"about_ca_system_score_gemma":0.001631083,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005801046,"about_ca_topic_score_gemma":0.01106849,"domain_scores_codex":[0.9994628,0.000080547,0.00002108764,0.0001701494,0.0001127229,0.0001527048],"domain_scores_gemma":[0.9991614,0.0003212555,0.00005586696,0.0002379481,0.0001590946,0.00006449458],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.001722469,0.0009438876,0.008070422,0.000496141,0.0005759089,0.0006699641,0.0002225338,0.1413152,0.04863956,0.005126044,0.09231929,0.6998986],"study_design_scores_gemma":[0.00003856216,0.0001597964,0.0004398669,0.00002753959,0.00004151039,0.00007179442,0.00002800274,0.9755964,0.01738001,0.003324938,0.002874235,0.00001737732],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2292754,0.002815861,0.5712873,0.002184738,0.00162786,0.0004209659,0.003488127,0.1767852,0.01211444],"genre_scores_gemma":[0.7309024,0.0004343326,0.2375219,0.002018909,0.0001704985,0.000335236,0.00667392,0.002972065,0.01897063],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.006001786,"threshold_uncertainty_score":0.020078,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02439590192211935,"score_gpt":0.3306339181913617,"score_spread":0.3062380162692424,"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."}}