{"id":"W4378942602","doi":"10.1145/3597926.3598135","title":"How Effective Are Neural Networks for Fixing Security Vulnerabilities","year":2023,"lang":"en","type":"preprint","venue":"","topic":"Software Testing and Debugging Techniques","field":"Computer Science","cited_by":90,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta; York University","funders":"National Science Foundation","keywords":"Computer science; Task (project management); Vulnerability (computing); Code (set theory); Software security assurance; Secure coding; Computer security; Source code; Software; Artificial neural network; Software bug; Automation; Programming language; Software engineering; Artificial intelligence; Information security; Engineering; Security service","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.0008726011,0.0003804102,0.0004805297,0.0001597974,0.0002324434,0.00105574,0.001366067,0.000350977,8.193198e-7],"category_scores_gemma":[0.001835534,0.0003427662,0.0002822644,0.0002115168,0.00006919151,0.0002087888,0.002401987,0.0007886444,0.000001501152],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001089059,"about_ca_system_score_gemma":0.00004347981,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002831731,"about_ca_topic_score_gemma":0.00003578175,"domain_scores_codex":[0.99783,0.0001847783,0.0002308338,0.00102829,0.0002370749,0.0004890027],"domain_scores_gemma":[0.9945963,0.003608647,0.0002430858,0.001194374,0.0002655447,0.00009203493],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00007131168,0.0002843323,0.04231357,0.00434546,0.0004618696,0.0001486389,0.006339779,0.1051589,0.000005788671,0.02755996,0.6546026,0.1587078],"study_design_scores_gemma":[0.00007181149,0.00005991511,0.0007383786,0.0002034637,0.000008915395,0.000004326102,0.00001342485,0.6826349,0.00006673793,0.3157921,0.00009767858,0.000308345],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.002840188,0.0002005022,0.9511022,0.002611427,0.001986894,0.001179431,0.00001729459,0.03998609,0.0000759638],"genre_scores_gemma":[0.8588312,0.000007329737,0.1383414,0.0002650172,0.0005549084,0.001341586,0.00002313561,0.0000493311,0.0005861693],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8559909,"threshold_uncertainty_score":0.9999813,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03806044719484671,"score_gpt":0.2861846731068167,"score_spread":0.24812422591197,"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."}}