{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch","bibliometrics"],"consensus_categories":[],"category_scores_codex":[0.0009531016,0.0008055745,0.0004801397,0.009848638,0.001248601,0.002782255,0.0003730898,0.0007881697,0.008189142],"category_scores_gemma":[0.008345485,0.0002222881,0.0003108279,0.01746688,0.001599964,0.005566981,0.001703399,0.0008407736,0.004422661],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007669192,"about_ca_system_score_gemma":0.0007492275,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003211865,"about_ca_topic_score_gemma":0.005509136,"domain_scores_codex":[0.9985548,0.00031956,0.0003220722,0.0002848404,0.0004631329,0.00005552548],"domain_scores_gemma":[0.9940262,0.003027087,0.001403603,0.0004415509,0.0008836435,0.0002179677],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.0007836614,0.0001482975,0.09401266,0.007702911,0.0005074582,0.002723788,0.02838877,0.001577211,0.0190203,0.09488308,0.2522082,0.4980437],"study_design_scores_gemma":[0.00004149637,0.000135113,0.1863835,0.001595921,0.0001996414,0.004658591,0.01528652,0.003708792,0.004502331,0.02772379,0.755583,0.0001812272],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5711943,0.03906748,0.03582389,0.01359357,0.004206842,0.0007260188,0.2051091,0.002327774,0.127951],"genre_scores_gemma":[0.8279544,0.01562109,0.05258653,0.001991472,0.001522654,0.0006805063,0.08012147,0.000694797,0.01882706],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9990469,"threshold_uncertainty_score":0.02739543,"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."}}