{"id":"W4403990800","doi":"10.1007/978-3-031-72986-7_25","title":"TIBET: Identifying and Evaluating Biases in Text-to-Image Generative Models","year":2024,"lang":"en","type":"book-chapter","venue":"Lecture notes in computer science","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":11,"is_retracted":false,"has_abstract":false,"ca_institutions":"Vector Institute; University of British Columbia","funders":"","keywords":"Computer science; Generative grammar; Image (mathematics); Artificial intelligence; Generative model; Natural language processing; Information retrieval","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"],"consensus_categories":[],"category_scores_codex":[0.001432174,0.0004782771,0.0004509625,0.001442359,0.0002504042,0.00125514,0.001989805,0.0001864499,0.00001052945],"category_scores_gemma":[0.0002415096,0.0004566625,0.00006838868,0.001085281,0.0003342305,0.0007216838,0.002280409,0.001066232,0.00009354665],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003247106,"about_ca_system_score_gemma":0.0003352869,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002431543,"about_ca_topic_score_gemma":0.0002138466,"domain_scores_codex":[0.9959447,0.00005731745,0.0005334938,0.002004765,0.0008509915,0.0006087492],"domain_scores_gemma":[0.9976466,0.0008082977,0.0001646042,0.001010147,0.000175396,0.0001949703],"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":[0.000002103689,0.00001577642,0.00004867623,0.00005303205,0.00000754553,0.00004938345,0.004334055,0.5341828,0.001759457,0.03302807,0.000006963407,0.4265122],"study_design_scores_gemma":[0.00009839448,0.00006276883,0.0002570979,0.0006105229,0.00000520674,0.00002539077,4.412075e-7,0.7826997,0.0003722278,0.2154562,0.00003466683,0.0003773772],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.003371716,0.0006528969,0.9903692,0.001848077,0.0004708188,0.0006293731,0.000006731352,0.0001902912,0.002460907],"genre_scores_gemma":[0.34317,0.00001876827,0.655461,0.0008028002,0.0002006937,0.00004450059,0.000003700284,0.00003966468,0.0002588737],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.4261348,"threshold_uncertainty_score":0.9997885,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0669570626640727,"score_gpt":0.353602100343215,"score_spread":0.2866450376791423,"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."}}