{"id":"W3196643997","doi":"10.18653/v1/2021.findings-emnlp.203","title":"Refining BERT Embeddings for Document Hashing via Mutual Information Maximization","year":2021,"lang":"en","type":"preprint","venue":"","topic":"Advanced Image and Video Retrieval Techniques","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal","funders":"Natural Science Foundation of Guangdong Province; National Natural Science Foundation of China","keywords":"Computer science; Hash function; Benchmark (surveying); Mutual information; Generative grammar; Generative model; Maximization; Artificial intelligence; Machine learning; Data mining; Theoretical computer science; Mathematics","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":[0.00206882,0.001254571,0.001522506,0.001355622,0.0005558258,0.001125663,0.001846024,0.001314083,0.002417442],"category_scores_gemma":[0.008001627,0.0006610482,0.0009619655,0.001848249,0.001533256,0.00352226,0.002438504,0.001885434,0.001783041],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001165656,"about_ca_system_score_gemma":0.00118484,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001375066,"about_ca_topic_score_gemma":0.001872464,"domain_scores_codex":[0.9983217,0.0006810984,0.0001023516,0.0003988768,0.0003854104,0.0001106461],"domain_scores_gemma":[0.9969055,0.001581299,0.0003214197,0.0006405333,0.0004169922,0.0001342161],"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":[0.0003675829,0.000240647,0.002786401,0.0004121463,0.000152509,0.0001835597,0.0003865659,0.3805581,0.01676976,0.0805079,0.009273548,0.5083613],"study_design_scores_gemma":[0.00001487384,0.00007342788,0.0002265823,0.00001130232,0.00001207689,0.0001155236,0.00002914943,0.9680603,0.002860999,0.02732804,0.001241521,0.00002632638],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.007911623,0.0002217995,0.9905529,0.00008954566,0.00002014204,0.00004176175,0.00008820894,0.0004258041,0.0006482965],"genre_scores_gemma":[0.364391,0.0006568127,0.6258647,0.0002791458,0.0002106013,0.000359431,0.001480153,0.0004371727,0.006320972],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.002417442,"threshold_uncertainty_score":0.01094109,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01598472026073924,"score_gpt":0.291538124228188,"score_spread":0.2755534039674488,"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."}}