{"id":"W2954460650","doi":"","title":"Empirical Analysis of Beam Search Performance Degradation in Neural Sequence Models","year":2019,"lang":"en","type":"article","venue":"International Conference on Machine Learning","topic":"Advanced Image and Video Retrieval Techniques","field":"Computer Science","cited_by":42,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Toronto","funders":"","keywords":"Degradation (telecommunications); Sequence (biology); Computer science; Artificial neural network; Artificial intelligence; Pattern recognition (psychology); Chemistry; Telecommunications","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":[0.0003531174,0.0001181331,0.0002156781,0.0006672557,0.00003218229,0.00007093512,0.000751545,0.00004605621,0.0001004246],"category_scores_gemma":[0.00006584687,0.000109272,0.00007423441,0.00088414,0.0000308273,0.000951172,0.0001759296,0.000434155,0.00001625582],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00008013279,"about_ca_system_score_gemma":0.00005174659,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000108882,"about_ca_topic_score_gemma":0.00001209929,"domain_scores_codex":[0.9986082,0.00008693388,0.0002895684,0.0003470299,0.0004976817,0.0001705888],"domain_scores_gemma":[0.9992535,0.0001140325,0.0001274196,0.0002373977,0.0002309621,0.00003674776],"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":[0.00007625914,0.00008331132,0.4262606,0.00001492651,0.00009020739,0.000008500459,0.0005409751,0.4363675,0.007794155,0.03130681,0.000002812227,0.09745401],"study_design_scores_gemma":[0.0001503561,0.000161996,0.0284534,0.00003476455,0.000006680079,0.000002004063,0.00001799447,0.965799,0.004493976,0.0007326402,0.00003989303,0.0001073079],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6629065,0.00001825951,0.3326529,0.0004224247,0.00005755664,0.00009021795,0.000003723158,0.00006729786,0.003781112],"genre_scores_gemma":[0.9932989,0.0000830592,0.005991126,0.0001266681,0.000008875547,0.000005286025,0.00003804932,0.000005689638,0.0004422965],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5294315,"threshold_uncertainty_score":0.4455983,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1093036857200191,"score_gpt":0.3857356482147963,"score_spread":0.2764319624947772,"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."}}