{"id":"W4400477472","doi":"10.2139/ssrn.4889071","title":"How to Tame Pre-Trained Transformers for Authorship Verification","year":2024,"lang":"en","type":"preprint","venue":"SSRN Electronic Journal","topic":"Authorship Attribution and Profiling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":false,"ca_institutions":"McGill University","funders":"","keywords":"Transformer; Computer science; Engineering; Electrical 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":["metaepi_narrow","scholarly_communication","research_integrity"],"consensus_categories":[],"category_scores_codex":[0.005537735,0.0003917508,0.0003727344,0.0004056657,0.0003060466,0.00121757,0.001439522,0.0004726514,0.000004304822],"category_scores_gemma":[0.0002372593,0.000365189,0.0004577252,0.0004617702,0.00002950295,0.0002641953,0.0002706833,0.005426239,0.00003288278],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001685017,"about_ca_system_score_gemma":0.005383132,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001071026,"about_ca_topic_score_gemma":0.00009275408,"domain_scores_codex":[0.9953948,0.0001982755,0.0004810637,0.0008000318,0.0004836032,0.00264221],"domain_scores_gemma":[0.9986798,0.00009511836,0.0002472719,0.0004110771,0.0002465556,0.0003202236],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00009011631,0.00003234072,0.000008416627,0.0001740231,0.0002408808,0.000002695786,0.002259729,0.0002463521,0.0009336657,0.8248481,0.0004018703,0.1707618],"study_design_scores_gemma":[0.0003181296,0.000374298,0.0000450296,0.0001600213,0.0000841893,0.0001312776,0.0003382717,0.01979326,0.001452865,0.9665616,0.01021429,0.0005268087],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.00549329,0.002201659,0.9420065,0.04667476,0.002225117,0.0009478585,0.00003108439,0.0002750566,0.000144642],"genre_scores_gemma":[0.9801463,0.0005246201,0.01159698,0.0002428131,0.0008480631,0.0001833098,0.00004695856,0.00005337322,0.006357527],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9746531,"threshold_uncertainty_score":0.99988,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02790066795711963,"score_gpt":0.3012856244737458,"score_spread":0.2733849565166261,"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."}}