{"id":"W4405140008","doi":"10.1007/978-981-96-0917-8_17","title":"Vision Language Models are blind","year":2024,"lang":"en","type":"book-chapter","venue":"Lecture notes in computer science","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":28,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Alberta","funders":"","keywords":"Computer science; Artificial intelligence; Natural language processing; Computer vision; Speech recognition","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"],"consensus_categories":[],"category_scores_codex":[0.0007295493,0.000504068,0.000434048,0.000995828,0.0002253285,0.000910163,0.003754475,0.0003193092,0.00002577535],"category_scores_gemma":[0.00005324244,0.0004492687,0.0001462222,0.000815464,0.0003806997,0.0005995815,0.002086716,0.001387958,0.0004050672],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002774372,"about_ca_system_score_gemma":0.0002874842,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001149689,"about_ca_topic_score_gemma":0.00009711508,"domain_scores_codex":[0.9960784,0.00002457404,0.0004335933,0.001884384,0.0009742976,0.0006047406],"domain_scores_gemma":[0.9971272,0.0003537417,0.0002354975,0.001925973,0.0001547539,0.000202834],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00000423005,0.00003966134,0.00002148379,0.00007876741,0.00001321316,0.0001597314,0.001743212,0.3316375,0.0002331035,0.1354331,0.00007433537,0.5305617],"study_design_scores_gemma":[0.0001133303,0.00004774601,0.00006643742,0.0003290191,0.000005638938,0.00003364165,1.450946e-7,0.7331767,0.0001214533,0.2651761,0.0005598975,0.0003698994],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0002187004,0.000853511,0.9831316,0.003025596,0.0009687793,0.0004114015,0.000008788063,0.0005002334,0.01088132],"genre_scores_gemma":[0.5044688,0.00003356642,0.4915295,0.00130672,0.0006051253,0.00002691232,0.00001020621,0.00007512335,0.001944083],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.5301918,"threshold_uncertainty_score":0.9997959,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01704194164196904,"score_gpt":0.2930850625544102,"score_spread":0.2760431209124412,"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."}}