{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/neural-variational-inference-and-learning-in-1","title":"Neural Variational Inference and Learning in Belief Networks","arxiv_id":"1402.0030","date":"2014-01-31","proceeding":null,"authors":["Andriy Mnih","Karol Gregor"],"abstract":"Highly expressive directed latent variable models, such as sigmoid belief\nnetworks, are difficult to train on large datasets because exact inference in\nthem is intractable and none of the approximate inference methods that have\nbeen applied to them scale well. We propose a fast non-iterative approximate\ninference method that uses a feedforward network to implement efficient exact\nsampling from the variational posterior. The model and this inference network\nare trained jointly by maximizing a variational lower bound on the\nlog-likelihood. Although the naive estimator of the inference model gradient is\ntoo high-variance to be useful, we make it practical by applying several\nstraightforward model-independent variance reduction techniques. Applying our\napproach to training sigmoid belief networks and deep autoregressive networks,\nwe show that it outperforms the wake-sleep algorithm on MNIST and achieves\nstate-of-the-art results on the Reuters RCV1 document dataset.","url_abs":"http://arxiv.org/abs/1402.0030v2","url_pdf":"http://arxiv.org/pdf/1402.0030v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"neural-variational-inference-and-learning-in-1","repo_url":"https://github.com/tensorflow/models","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"neural-variational-inference-and-learning-in-1","repo_url":"https://github.com/tensorflow/models/tree/master/research/rebar","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"variational-inference","task_name":"Variational Inference"}],"methods":[{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"feedforward-network","method_name":"Feedforward Network"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1402.0030","atlas_url":"https://app.syntology.ai/?focus=1402.0030","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}