{"id":"W4293868331","doi":"10.1109/crv55824.2022.00015","title":"Semi-supervised Grounding Alignment for Multi-modal Feature Learning","year":2022,"lang":"en","type":"article","venue":"","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"Vector Institute; Canadian Institute for Advanced Research; University of British Columbia","funders":"","keywords":"Computer science; Artificial intelligence; Phrase; Margin (machine learning); Leverage (statistics); Machine learning; Ground; Sentence; Feature (linguistics); Transformer; Modal; Classifier (UML); Feature learning; Natural language processing; Question answering; Supervised learning; Feature extraction; Pattern recognition (psychology); Artificial neural network; Engineering","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":[0.002041986,0.001770236,0.001269794,0.00114315,0.0005787144,0.001128678,0.003816073,0.002170824,0.004801504],"category_scores_gemma":[0.005592154,0.0005893369,0.001113985,0.001423507,0.001557107,0.004359047,0.002975153,0.003140848,0.002320959],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001386455,"about_ca_system_score_gemma":0.0009716297,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002293368,"about_ca_topic_score_gemma":0.00376164,"domain_scores_codex":[0.998257,0.0004911006,0.00007080793,0.0007309109,0.0003059199,0.0001442237],"domain_scores_gemma":[0.9975636,0.0009129492,0.0002792729,0.0008015276,0.0003446152,0.00009813524],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0003707969,0.000429045,0.001487881,0.0002288465,0.0001207667,0.0001340524,0.0002263621,0.2391682,0.02439733,0.01816182,0.01561435,0.6996606],"study_design_scores_gemma":[0.00001334738,0.00008072792,0.0002187544,0.000009131072,0.000008057435,0.00003754491,0.00002205325,0.9756221,0.004936863,0.01795205,0.001088491,0.0000109304],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0109434,0.0001536175,0.9832138,0.0001457687,0.00002352729,0.00006660527,0.00026441,0.004414237,0.0007746669],"genre_scores_gemma":[0.5182625,0.000186623,0.4706059,0.0005684337,0.0001408971,0.0004811917,0.00439473,0.0007600642,0.004599716],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004801504,"threshold_uncertainty_score":0.01606268,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03163132250895909,"score_gpt":0.3021625355299601,"score_spread":0.270531213021001,"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."}}