{"id":"W4225853051","doi":"10.5281/zenodo.6368376","title":"DP-VAE: Human-Readable Text Anonymization for Online Reviews with Differentially Private Variational Autoencoders","year":2022,"lang":"en","type":"article","venue":"Zenodo (CERN European Organization for Nuclear Research)","topic":"Authorship Attribution and Profiling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Artificial intelligence; Private information retrieval; Data science; Information retrieval; Computer security","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":[],"category_scores_codex":[0.001227946,0.0001563569,0.0001844973,0.0002082679,0.004437072,0.0006463565,0.00146562,0.00003842802,0.005199574],"category_scores_gemma":[0.0002565559,0.0001530268,0.00006096192,0.0007575082,0.0000585956,0.0004295065,0.001232423,0.0002725192,0.000386901],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002095391,"about_ca_system_score_gemma":0.00001573587,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000006748106,"about_ca_topic_score_gemma":6.499948e-7,"domain_scores_codex":[0.9977086,0.000566753,0.0003521572,0.000516112,0.0004985473,0.0003578367],"domain_scores_gemma":[0.9987051,0.00003735443,0.0002471071,0.0004802593,0.000402021,0.0001281613],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0002375763,0.0009498082,0.00003664422,0.0002487563,0.0001299931,0.00001269654,0.002770626,0.01147646,0.00851322,0.7052523,0.1317809,0.1385911],"study_design_scores_gemma":[0.0007361128,0.000397823,0.0004968879,0.00001837914,0.00001408259,0.00003959138,0.00004567131,0.04813999,0.000180926,0.00151192,0.9481979,0.0002207271],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.00159926,0.00008109048,0.9913952,0.00180198,0.0001338414,0.0009084666,0.0002511354,0.0006753478,0.003153639],"genre_scores_gemma":[0.672133,0.0002299154,0.2666265,0.003000795,0.0008202957,0.000007048071,0.0408659,0.003107331,0.01320917],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.816417,"threshold_uncertainty_score":0.996859,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06481148837151597,"score_gpt":0.2794793357719953,"score_spread":0.2146678474004793,"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."}}