{"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/iterative-refinement-of-the-approximate","title":"Iterative Refinement of the Approximate Posterior for Directed Belief Networks","arxiv_id":"1511.06382","date":"2015-11-19","proceeding":"NeurIPS 2016 12","authors":["R. Devon Hjelm","Kyunghyun Cho","Junyoung Chung","Russ Salakhutdinov","Vince Calhoun","Nebojsa Jojic"],"abstract":"Variational methods that rely on a recognition network to approximate the\nposterior of directed graphical models offer better inference and learning than\nprevious methods. Recent advances that exploit the capacity and flexibility in\nthis approach have expanded what kinds of models can be trained. However, as a\nproposal for the posterior, the capacity of the recognition network is limited,\nwhich can constrain the representational power of the generative model and\nincrease the variance of Monte Carlo estimates. To address these issues, we\nintroduce an iterative refinement procedure for improving the approximate\nposterior of the recognition network and show that training with the refined\nposterior is competitive with state-of-the-art methods. The advantages of\nrefinement are further evident in an increased effective sample size, which\nimplies a lower variance of gradient estimates.","url_abs":"http://arxiv.org/abs/1511.06382v6","url_pdf":"http://arxiv.org/pdf/1511.06382v6.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":"iterative-refinement-of-the-approximate","repo_url":"https://github.com/rdevon/IRVI","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}