{"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/asymmetric-variational-autoencoders","title":"Asymmetric Variational Autoencoders","arxiv_id":"1711.08352","date":"2017-11-20","proceeding":null,"authors":["Guoqing Zheng","Yiming Yang","Jaime Carbonell"],"abstract":"Variational inference for latent variable models is prevalent in various\nmachine learning problems, typically solved by maximizing the Evidence Lower\nBound (ELBO) of the true data likelihood with respect to a variational\ndistribution. However, freely enriching the family of variational distribution\nis challenging since the ELBO requires variational likelihood evaluations of\nthe latent variables. In this paper, we propose a novel framework to enrich the\nvariational family by incorporating auxiliary variables to the variational\nfamily. The resulting inference network doesn't require density evaluations for\nthe auxiliary variables and thus complex implicit densities over the auxiliary\nvariables can be constructed by neural networks. It can be shown that the\nactual variational posterior of the proposed approach is essentially modeling a\nrich probabilistic mixture of simple variational posterior indexed by auxiliary\nvariables, thus a flexible inference model can be built. Empirical evaluations\non several density estimation tasks demonstrates the effectiveness of the\nproposed method.","url_abs":"http://arxiv.org/abs/1711.08352v2","url_pdf":"http://arxiv.org/pdf/1711.08352v2.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":"asymmetric-variational-autoencoders","repo_url":"https://github.com/apsyx/mvae","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"density-estimation","task_name":"Density Estimation"},{"task_slug":"variational-inference","task_name":"Variational Inference"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}