{"id":"W4386566652","doi":"10.18653/v1/2023.eacl-main.195","title":"On Evaluation of Document Classification with RVL-CDIP","year":2023,"lang":"en","type":"article","venue":"","topic":"Imbalanced Data Classification Techniques","field":"Computer Science","cited_by":13,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Atomic Energy of Canada Limited","keywords":"Benchmarking; Benchmark (surveying); Computer science; Scope (computer science); Resource (disambiguation); Artificial intelligence; Task (project management); Machine learning; Information retrieval; Data mining; Pattern recognition (psychology); Engineering; Business","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":[0.02651795,0.002624899,0.002133436,0.006240978,0.00193673,0.005232151,0.00446333,0.003471538,0.003393381],"category_scores_gemma":[0.06128775,0.0005306094,0.001241702,0.005650853,0.001472634,0.00631924,0.003512573,0.003142198,0.004956219],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.004514728,"about_ca_system_score_gemma":0.002173487,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0183849,"about_ca_topic_score_gemma":0.01559372,"domain_scores_codex":[0.9789503,0.008967046,0.0020434,0.00315976,0.005877822,0.001001738],"domain_scores_gemma":[0.9587022,0.02303218,0.001709907,0.005821413,0.009271001,0.001463187],"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":[0.003070158,0.002175259,0.02258937,0.001616431,0.000750228,0.0002987691,0.0004755217,0.1259488,0.007348218,0.004557563,0.1242085,0.7069612],"study_design_scores_gemma":[0.0003600208,0.001966516,0.01182246,0.0003502606,0.0001422637,0.0003910576,0.0006990383,0.9369278,0.01752234,0.005081187,0.02461359,0.0001233743],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6217164,0.03636609,0.1897594,0.007389816,0.005944955,0.002388116,0.02626696,0.05727662,0.05289173],"genre_scores_gemma":[0.7680435,0.002530792,0.1619957,0.001709222,0.0006994391,0.0007187899,0.05300654,0.001560409,0.009735713],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.02651795,"threshold_uncertainty_score":0.140242,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07528893480864146,"score_gpt":0.3406554908920777,"score_spread":0.2653665560834362,"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."}}