{"id":"W3174481471","doi":"10.18653/v1/2021.acl-long.564","title":"Mind Your Outliers! Investigating the Negative Impact of Outliers on Active Learning for Visual Question Answering","year":2021,"lang":"en","type":"article","venue":"","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":50,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Open Philanthropy Project; Canadian Institute for Advanced Research","keywords":"Question answering; Outlier; Computational linguistics; Natural language processing; Computer science; Artificial intelligence; Corpus linguistics; Volume (thermodynamics); Linguistics; Information retrieval; Philosophy","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.009100189,0.0007419929,0.001108096,0.0007864059,0.001037657,0.00228407,0.002051712,0.00240273,0.003194824],"category_scores_gemma":[0.09884331,0.0004556161,0.0004790663,0.000796925,0.001174225,0.005813006,0.002444939,0.003218982,0.00159588],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005616857,"about_ca_system_score_gemma":0.0004885698,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003446736,"about_ca_topic_score_gemma":0.003642747,"domain_scores_codex":[0.9948037,0.002813588,0.0001697957,0.0008306495,0.001191763,0.000190503],"domain_scores_gemma":[0.9137367,0.07272369,0.002530921,0.004477841,0.005426312,0.001104488],"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":[0.005777064,0.001510745,0.05005126,0.001015926,0.0004802217,0.0004836232,0.003192306,0.05416472,0.02827407,0.01840862,0.09949065,0.7371508],"study_design_scores_gemma":[0.0002106069,0.0008374171,0.01650516,0.0001857054,0.0001592907,0.0003766571,0.001788008,0.8824614,0.01736771,0.06480104,0.01523214,0.00007497898],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6390908,0.01693474,0.303103,0.01672932,0.002023384,0.0002780527,0.002756688,0.005689125,0.01339498],"genre_scores_gemma":[0.9548419,0.0008501634,0.0375229,0.0009493959,0.0004948562,0.00009593361,0.00169161,0.0003669607,0.003186277],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.009100189,"threshold_uncertainty_score":0.04812694,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02975810514562487,"score_gpt":0.3674712042243422,"score_spread":0.3377130990787173,"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."}}