{"id":"W4388954848","doi":"10.1145/3605770.3625215","title":"Distinguishing AI- and Human-Generated Code: A Case Study","year":2023,"lang":"en","type":"article","venue":"","topic":"Software Engineering Research","field":"Computer Science","cited_by":17,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Calgary","funders":"","keywords":"Computer science; Code (set theory); Task (project management); Code review; Software; Source code; Artificial intelligence; Programming language; Natural language processing; Software engineering; Computer security; Static program analysis; Software development; Engineering","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":[],"consensus_categories":[],"category_scores_codex":[0.0004901154,0.00007530612,0.00008169098,0.0001355774,0.0001865174,0.0003466959,0.0002917181,0.00002026449,0.000006764246],"category_scores_gemma":[0.0004916605,0.00006740702,0.00001046622,0.0008351826,0.0000148586,0.0001547609,0.0004887987,0.0001271305,0.00003387978],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001811093,"about_ca_system_score_gemma":0.00001940719,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0006969896,"about_ca_topic_score_gemma":0.0002031752,"domain_scores_codex":[0.9991168,0.00003747713,0.0001028962,0.0002948935,0.000212408,0.0002355001],"domain_scores_gemma":[0.9992015,0.0002923754,0.000009805766,0.000341147,0.00006616069,0.00008896798],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00000308962,0.0003630751,0.8247947,0.00009264266,0.0001244305,0.08459708,0.01595499,0.003147859,0.002822261,0.0124141,0.02391018,0.03177561],"study_design_scores_gemma":[0.001442089,0.0005425237,0.1904128,0.00002869933,0.00001002003,0.002456168,0.001599626,0.8005038,0.0008151099,0.0004495502,0.001031006,0.0007086646],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9506174,0.00000997675,0.0476109,0.0001460546,0.0001116668,0.0001478495,7.115565e-7,0.001258489,0.00009693095],"genre_scores_gemma":[0.9967591,3.748106e-7,0.002335953,0.00003094467,0.00003237169,0.00001822759,9.062333e-7,0.000009408811,0.0008127387],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.7973559,"threshold_uncertainty_score":0.3343196,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06023643670637761,"score_gpt":0.3549796554843336,"score_spread":0.294743218777956,"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."}}