{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0007945563,0.0006682178,0.0007486577,0.001100496,0.0006285821,0.0009948636,0.001389812,0.001429572,0.008518678],"category_scores_gemma":[0.005959865,0.0004129341,0.001106127,0.001088499,0.000953297,0.003476007,0.001796124,0.001787291,0.001188076],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001078286,"about_ca_system_score_gemma":0.0005430257,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006209723,"about_ca_topic_score_gemma":0.008733471,"domain_scores_codex":[0.9995555,0.0001758691,0.00002030294,0.0001637148,0.00004958332,0.00003509552],"domain_scores_gemma":[0.9985233,0.0009367147,0.00008994652,0.0002606402,0.0001247239,0.00006463176],"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":[0.0004212491,0.0002690482,0.001747053,0.0004986876,0.0001687689,0.000248297,0.0008304032,0.239314,0.01832343,0.3059934,0.01472682,0.4174589],"study_design_scores_gemma":[0.0000117238,0.00002173788,0.0002378069,0.00001963528,0.00002178283,0.00002292134,0.00004597138,0.7163628,0.001276218,0.2803155,0.001654338,0.000009546802],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.03881907,0.0007361886,0.9526677,0.0008728802,0.00007138771,0.0001078349,0.0005721598,0.001739044,0.004413774],"genre_scores_gemma":[0.7101814,0.0007466704,0.2779891,0.0004964327,0.0001396185,0.0002837936,0.001798684,0.000424049,0.00794048],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.008518678,"threshold_uncertainty_score":0.02849776,"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."}}