{"id":"W4285490440","doi":"10.1145/3533767.3534220","title":"DocTer: documentation-guided fuzzing for testing deep learning API functions","year":2022,"lang":"en","type":"preprint","venue":"","topic":"Software Engineering Research","field":"Computer Science","cited_by":73,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"National Science Foundation","keywords":"Fuzz testing; Computer science; Documentation; Parsing; Function (biology); Constraint (computer-aided design); Software bug; Artificial intelligence; Dependency (UML); Software; Programming language; Machine learning; Data mining","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":[0.004817409,0.002360279,0.0007966589,0.002293271,0.0007586462,0.002085852,0.003953923,0.00208953,0.01861915],"category_scores_gemma":[0.04420583,0.001492822,0.001632549,0.001088208,0.002107423,0.004932171,0.003255012,0.002678775,0.004354498],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001540007,"about_ca_system_score_gemma":0.002800844,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003660599,"about_ca_topic_score_gemma":0.008108934,"domain_scores_codex":[0.992991,0.002302941,0.0007021911,0.001240081,0.002260836,0.0005029727],"domain_scores_gemma":[0.9694574,0.01961631,0.001250629,0.00691068,0.002368097,0.0003968173],"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":[0.002000425,0.0008732858,0.01963851,0.002455078,0.0003371814,0.001457592,0.0008702662,0.1262358,0.04864303,0.04989602,0.1292096,0.6183832],"study_design_scores_gemma":[0.0004252612,0.0003076564,0.00253578,0.0003361024,0.00009142308,0.0007046236,0.000197902,0.7960191,0.1024132,0.0586867,0.03815477,0.0001274242],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.05133399,0.0005268404,0.7654034,0.0008404351,0.0002504396,0.0004085841,0.005575373,0.1676906,0.007970275],"genre_scores_gemma":[0.37608,0.0003000507,0.5870193,0.0008296258,0.00007397209,0.0006709633,0.01004333,0.01876635,0.006216466],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01861915,"threshold_uncertainty_score":0.06228727,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06039005306412008,"score_gpt":0.331886397911752,"score_spread":0.2714963448476319,"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."}}