{"id":"W4324297696","doi":"10.48550/arxiv.2303.06992","title":"Improving Mutual Information Estimation with Annealed and Energy-Based Bounds","year":2023,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Generative Adversarial Networks and Image Synthesis","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Canadian Institute for Advanced Research","keywords":"Mutual information; Marginal distribution; Upper and lower bounds; Joint probability distribution; Computer science; Sample size determination; Density estimation; Sampling (signal processing); Joint (building); Partition (number theory); Algorithm; Sample (material); Mathematics; Applied mathematics; Mathematical optimization; Statistics; Artificial intelligence; Combinatorics; Random variable; Estimator; Mathematical analysis","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":[0.005875493,0.002224295,0.002273472,0.002526582,0.0008344083,0.002459674,0.003984964,0.002618392,0.003458004],"category_scores_gemma":[0.0506033,0.00138114,0.001661217,0.001570459,0.002637269,0.006885835,0.005415685,0.004637932,0.0008381295],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002621162,"about_ca_system_score_gemma":0.002221143,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003752241,"about_ca_topic_score_gemma":0.005008128,"domain_scores_codex":[0.995759,0.001373854,0.0002305672,0.0006962351,0.001592913,0.0003475026],"domain_scores_gemma":[0.973677,0.02060481,0.001229751,0.002425687,0.001568566,0.0004941174],"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":[0.0002784105,0.0001598412,0.002971779,0.0002722022,0.0001575086,0.0001925652,0.0003531502,0.7503587,0.005060114,0.1448062,0.003386256,0.09200326],"study_design_scores_gemma":[0.000009543041,0.00002493461,0.0002076084,0.00002229857,0.00001543145,0.00003378411,0.00001071217,0.9697583,0.001346451,0.02797229,0.0005827709,0.00001581466],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01222569,0.0008722126,0.9828942,0.0003388791,0.00005007146,0.00004363937,0.00008667378,0.0006613086,0.002827191],"genre_scores_gemma":[0.5851499,0.001010488,0.4060428,0.0007541937,0.0002863778,0.0003324474,0.0007186878,0.0007812909,0.004923919],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.005875493,"threshold_uncertainty_score":0.03107297,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03365478930670212,"score_gpt":0.1588642217671323,"score_spread":0.1252094324604302,"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."}}