{"id":"W6912806176","doi":"10.5281/zenodo.6583266","title":"Learning to Generate Inversion-Resistant Model Explanations","year":2022,"lang":"en","type":"article","venue":"Zenodo (CERN European Organization for Nuclear Research)","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Kootenay Association for Science & Technology","funders":"","keywords":"Inversion (geology); Rendering (computer graphics); Artificial neural network; Adversary; Deep neural networks; Threat model; Minimax; Adversarial system","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":["sts","insufficient_payload"],"consensus_categories":["insufficient_payload"],"category_scores_codex":[0.0009326749,0.0001235665,0.0001128323,0.0003806841,0.006996531,0.0007056024,0.002260139,0.00002527061,0.002754972],"category_scores_gemma":[0.0007458255,0.0001485385,0.00004438129,0.001129967,0.00003768404,0.0003954425,0.005614073,0.0005279686,0.001571909],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003168846,"about_ca_system_score_gemma":0.00001048745,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001489852,"about_ca_topic_score_gemma":1.55533e-7,"domain_scores_codex":[0.997716,0.000603863,0.0002058012,0.0005065992,0.00060852,0.0003592048],"domain_scores_gemma":[0.9989102,0.00004027305,0.00009710367,0.0004890678,0.0002709829,0.0001923163],"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":[0.00002540865,0.00004592556,0.000001502882,0.000006912612,0.00001152697,0.00001657999,0.003799181,0.8864577,0.004390495,0.03126685,0.04386143,0.03011643],"study_design_scores_gemma":[0.0001918492,0.0001488803,0.00001692864,0.000003408344,0.00000312999,0.00001579365,0.0004094208,0.5321701,0.00007768402,0.0003228951,0.466502,0.000137903],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01380724,0.00001773522,0.9492365,0.003089293,0.0001982018,0.0003153782,0.00004344375,0.001449993,0.03184224],"genre_scores_gemma":[0.9782435,0.000006364815,0.01782371,0.0005907436,0.00008456794,2.214247e-7,0.0003750606,0.0006170322,0.002258808],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9644362,"threshold_uncertainty_score":0.9992055,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03594158688028365,"score_gpt":0.2502878349863321,"score_spread":0.2143462481060484,"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."}}