{"id":"W4402727410","doi":"10.1109/cvpr52733.2024.01307","title":"Contrasting Intra-Modal and Ranking Cross-Modal Hard Negatives to Enhance Visio-Linguistic Compositional Understanding","year":2024,"lang":"en","type":"article","venue":"","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":8,"is_retracted":false,"has_abstract":true,"ca_institutions":"Mila - Quebec Artificial Intelligence Institute","funders":"","keywords":"Modal; Ranking (information retrieval); Computer science; Linguistics; Natural language processing; Artificial intelligence; Materials science; Philosophy","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":["scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.0004059674,0.0001938204,0.0001830564,0.0001740158,0.0004880864,0.00134629,0.0004114691,0.00005123838,0.00004182502],"category_scores_gemma":[0.0002682894,0.0001860621,0.0000476412,0.0004412854,0.0001123128,0.0003528895,0.0002683495,0.000317775,0.00007327583],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001599,"about_ca_system_score_gemma":0.00007320373,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001265838,"about_ca_topic_score_gemma":0.00001031523,"domain_scores_codex":[0.998295,0.00006479244,0.0002913297,0.0007030421,0.0003064095,0.0003393926],"domain_scores_gemma":[0.9984173,0.001036538,0.00004707832,0.0002532396,0.0000787255,0.0001670987],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00001805817,0.00003069932,0.003160357,0.00007293487,0.00004605922,0.00002251701,0.002759141,0.003492897,0.02724063,0.9511735,0.0001224343,0.01186072],"study_design_scores_gemma":[0.0001934206,0.00007908838,0.02156574,0.0002244103,0.000009952527,0.0000408886,0.00004454844,0.9319856,0.002648472,0.04257637,0.0003023169,0.0003291509],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1669397,0.0001397963,0.8263658,0.002221588,0.0002607062,0.0002182099,0.000007044748,0.0004953339,0.003351822],"genre_scores_gemma":[0.8568957,0.000003370621,0.1424507,0.0003352837,0.0001976837,0.00002970543,0.000004563753,0.00001625335,0.00006673009],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9284927,"threshold_uncertainty_score":0.9996904,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02464168471662963,"score_gpt":0.3549871254119355,"score_spread":0.3303454406953059,"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."}}