{"id":"W4220816919","doi":"10.1126/sciadv.abm2463","title":"Based on billions of words on the internet, <scp>people</scp> = <scp>men</scp>","year":2022,"lang":"en","type":"article","venue":"Science Advances","topic":"Authorship Attribution and Profiling","field":"Computer Science","cited_by":53,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"York University","keywords":"Similarity (geometry); Context (archaeology); Measure (data warehouse); The Internet; Linguistics; Computer science; Psychology; Similarity measure; Social psychology; Natural language processing; Sociology; World Wide Web; Artificial intelligence; History","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":[],"consensus_categories":[],"category_scores_codex":[0.00274411,0.0001922938,0.0001978764,0.0003768955,0.001039841,0.0001780962,0.003174092,0.00002994758,0.00003856389],"category_scores_gemma":[0.001773738,0.0001345331,0.0001043321,0.004008009,0.0005240139,0.0007100282,0.0006268037,0.0004247681,0.00006470218],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001283026,"about_ca_system_score_gemma":0.00034579,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001568851,"about_ca_topic_score_gemma":0.000009859368,"domain_scores_codex":[0.9967327,0.000258245,0.0003176921,0.0006980012,0.001404262,0.0005891206],"domain_scores_gemma":[0.9963141,0.002281462,0.0002803182,0.0008460188,0.0001260741,0.0001520357],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00001518009,0.0007792441,0.01806073,0.00005380275,0.00002207028,0.00002594748,0.01733627,0.1066281,0.01406526,0.7897803,0.01125558,0.04197757],"study_design_scores_gemma":[0.0007810757,0.002175279,0.01254502,0.0001697897,0.00001615127,0.00002056885,0.01145414,0.3841404,0.2808532,0.02815505,0.2794432,0.0002461158],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5923707,0.0009301143,0.3312088,0.006800101,0.004569426,0.001052725,0.00007340944,0.0005304521,0.0624642],"genre_scores_gemma":[0.993859,0.00001107618,0.002704908,0.001462139,0.00004097851,0.00007535,0.000002798907,0.00000811503,0.001835653],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.7616252,"threshold_uncertainty_score":0.7997724,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02511934123538057,"score_gpt":0.2815690368670433,"score_spread":0.2564496956316628,"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."}}