{"id":"W4413147826","doi":"10.1109/cvpr52734.2025.01361","title":"Assessing and Learning Alignment of Unimodal Vision and Language Models","year":2025,"lang":"en","type":"article","venue":"","topic":"Categorization, perception, and language","field":"Psychology","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Mila - Quebec Artificial Intelligence Institute","funders":"","keywords":"Computer science; Artificial intelligence; Computer vision; Natural language processing","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":[],"consensus_categories":[],"category_scores_codex":[0.0001719181,0.00006153918,0.00009735803,0.00008101926,0.00007341571,0.00003714086,0.00002721037,0.00005378789,0.0006262435],"category_scores_gemma":[0.000009106345,0.00005222251,0.00001351836,0.00007842105,0.0000422831,0.0001036879,0.00003026532,0.00006306884,0.000002799766],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000009782716,"about_ca_system_score_gemma":0.000008263241,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0007007488,"about_ca_topic_score_gemma":0.00003926965,"domain_scores_codex":[0.9995126,0.00007581327,0.0001201685,0.0001530879,0.00005572631,0.0000825686],"domain_scores_gemma":[0.9997832,0.00005196478,0.00003114368,0.00009062672,0.00001996074,0.00002314261],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"observational","study_design_scores_codex":[0.0001119011,0.0003559581,0.06041314,0.0002522614,0.0001817612,0.00001775092,0.1366719,0.0008178618,0.08305244,0.3946421,0.002536855,0.3209461],"study_design_scores_gemma":[0.003761067,0.0003071433,0.733511,0.0001925653,0.0001676223,0.00003321593,0.2262263,0.02223124,0.001958013,0.005915764,0.005137747,0.0005583449],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8940337,0.0005417551,0.03015863,0.0001084781,0.00006662837,0.00006305238,7.703883e-7,0.00003289804,0.07499412],"genre_scores_gemma":[0.9880026,0.00003582258,0.0004687214,0.00009480746,0.00001824053,0.000002490447,0.00001112715,0.000005476079,0.01136073],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.6730979,"threshold_uncertainty_score":0.6856927,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01402788751029604,"score_gpt":0.3499927015957561,"score_spread":0.33596481408546,"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."}}