{"id":"W3105009590","doi":"","title":"Multimodal Graph Networks for Compositional Generalization in Visual Question Answering","year":2020,"lang":"en","type":"article","venue":"Neural Information Processing Systems","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":27,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Toronto","funders":"","keywords":"Computer science; Question answering; Generalization; Graph; Artificial intelligence; Graph theory; Natural language processing; Theoretical computer science; Mathematics; Combinatorics","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":[0.0002436444,0.0001416837,0.0001544991,0.0001564372,0.0002289354,0.0005634334,0.0003031222,0.00008312639,8.639394e-7],"category_scores_gemma":[0.0000595799,0.0001467811,0.00003914906,0.0006092894,0.00001819161,0.002914136,0.00005162337,0.0001623867,0.000008830812],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006068592,"about_ca_system_score_gemma":0.00004190025,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001365879,"about_ca_topic_score_gemma":0.000002447501,"domain_scores_codex":[0.998719,0.00007119551,0.0005527466,0.0002076048,0.0002378151,0.0002116232],"domain_scores_gemma":[0.9992313,0.00004981189,0.0003012712,0.0001076509,0.000220161,0.00008978086],"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":[0.00001434378,0.00001568653,0.0025598,0.0001585175,0.000002481793,2.080637e-7,0.0009362136,0.968623,0.0004968447,0.007415575,0.00006324927,0.01971409],"study_design_scores_gemma":[0.0004786679,0.00005175469,0.008364079,0.00005496059,0.000002484396,0.000007569316,0.00003443695,0.9903178,0.00006903364,0.00004267113,0.000420756,0.0001557496],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05405854,0.00006033892,0.9432809,0.001403233,0.0001956287,0.0005851852,0.00000426725,0.0003488459,0.00006308937],"genre_scores_gemma":[0.9822419,0.000001708852,0.0163575,0.0007849505,0.0001658184,0.0002305188,0.0002061108,0.000009052368,0.000002402036],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9281834,"threshold_uncertainty_score":0.598556,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0154267698766019,"score_gpt":0.2875687495257047,"score_spread":0.2721419796491028,"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."}}