{"id":"W4416366377","doi":"10.1109/tvcg.2025.3634791","title":"Probing the Visualization Literacy of Vision Language Models: The Good, the Bad, and the Ugly","year":2025,"lang":"en","type":"article","venue":"IEEE Transactions on Visualization and Computer Graphics","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Visualization; Chart; Correctness; Comprehension; Data visualization; Program comprehension; Code (set theory); Visual reasoning; Key (lock)","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.004210357,0.001460048,0.0006152839,0.001032251,0.0004117248,0.004607277,0.002431084,0.001890435,0.007154249],"category_scores_gemma":[0.04026228,0.0005644654,0.001464206,0.0006469597,0.001249274,0.006683005,0.002496233,0.002741734,0.002016767],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001939576,"about_ca_system_score_gemma":0.001455082,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008358584,"about_ca_topic_score_gemma":0.01042539,"domain_scores_codex":[0.9978236,0.0009641357,0.0001167934,0.0005854096,0.0003703611,0.0001397096],"domain_scores_gemma":[0.9845521,0.01135359,0.0005815234,0.002003753,0.001155685,0.0003532057],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.002912763,0.0008069312,0.03890239,0.003201066,0.0006548996,0.001052808,0.01016536,0.167024,0.04826593,0.05219267,0.09348761,0.5813336],"study_design_scores_gemma":[0.0002479214,0.0004803774,0.008297482,0.0003067628,0.0002063631,0.0003518619,0.001719347,0.8422827,0.02745724,0.0783746,0.04012674,0.0001486294],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4182671,0.002226149,0.4277467,0.005654756,0.000499098,0.0006920061,0.009168637,0.1028191,0.0329264],"genre_scores_gemma":[0.843784,0.0004557261,0.1412418,0.0007598286,0.00005931497,0.0003786411,0.006109722,0.003275753,0.003935247],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.008358584,"threshold_uncertainty_score":0.02393335,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.009481657295951497,"score_gpt":0.3001178442489346,"score_spread":0.290636186952983,"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."}}