{"id":"W4403321808","doi":"10.48550/arxiv.2409.05657","title":"Adversarial Attacks on Data Attribution","year":2024,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Canadian Institute for Advanced Research","keywords":"Adversarial system; Attribution; Computer security; Computer science; Authorship attribution; Data science; Artificial intelligence; Psychology; Social psychology","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","open_science","research_integrity","insufficient_payload"],"consensus_categories":["open_science"],"category_scores_codex":[0.0006931585,0.0004864525,0.0004237243,0.0003909897,0.0002462225,0.0003184028,0.006297122,0.0005233353,0.00006062062],"category_scores_gemma":[0.0001711747,0.0005583082,0.0002081169,0.0008287712,0.000123189,0.0005596774,0.02114256,0.002320265,0.001019618],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005030017,"about_ca_system_score_gemma":0.0004356803,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001601181,"about_ca_topic_score_gemma":0.00002276449,"domain_scores_codex":[0.9961039,0.0003039063,0.0002732155,0.00257689,0.0002359407,0.0005061922],"domain_scores_gemma":[0.9949499,0.0002214506,0.0002694543,0.004252289,0.0001179126,0.0001890123],"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":[0.00004577151,0.00005097487,0.0002089822,0.0000824379,0.0001199657,0.0006415442,0.00008904575,0.5832428,0.000005126153,0.4106302,0.003160615,0.001722534],"study_design_scores_gemma":[0.000453606,0.00005774929,0.0001893768,0.0002238312,0.0001336515,0.000005683538,0.0000200348,0.9308807,0.00001847742,0.06013874,0.00731373,0.0005644748],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0156894,0.00008116149,0.9713047,0.0007385474,0.005741781,0.0003508914,0.0001321192,0.0009179679,0.005043441],"genre_scores_gemma":[0.993457,0.00006350352,0.003399214,0.000129501,0.0007097014,7.131031e-7,0.0003088386,0.00004025883,0.001891323],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9777675,"threshold_uncertainty_score":0.9999814,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1544047434716,"score_gpt":0.2447835755478195,"score_spread":0.09037883207621955,"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."}}