{"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/dropout-inference-in-bayesian-neural-networks","title":"Dropout Inference in Bayesian Neural Networks with Alpha-divergences","arxiv_id":"1703.02914","date":"2017-03-08","proceeding":"ICML 2017 8","authors":["Yingzhen Li","Yarin Gal"],"abstract":"To obtain uncertainty estimates with real-world Bayesian deep learning\nmodels, practical inference approximations are needed. Dropout variational\ninference (VI) for example has been used for machine vision and medical\napplications, but VI can severely underestimates model uncertainty.\nAlpha-divergences are alternative divergences to VI's KL objective, which are\nable to avoid VI's uncertainty underestimation. But these are hard to use in\npractice: existing techniques can only use Gaussian approximating\ndistributions, and require existing models to be changed radically, thus are of\nlimited use for practitioners. We propose a re-parametrisation of the\nalpha-divergence objectives, deriving a simple inference technique which,\ntogether with dropout, can be easily implemented with existing models by simply\nchanging the loss of the model. We demonstrate improved uncertainty estimates\nand accuracy compared to VI in dropout networks. We study our model's epistemic\nuncertainty far away from the data using adversarial images, showing that these\ncan be distinguished from non-adversarial images by examining our model's\nuncertainty.","url_abs":"http://arxiv.org/abs/1703.02914v1","url_pdf":"http://arxiv.org/pdf/1703.02914v1.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":"dropout-inference-in-bayesian-neural-networks","repo_url":"https://github.com/janisgp/Sampling-free-Epistemic-Uncertainty","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}}],"tasks":[{"task_slug":"variational-inference","task_name":"Variational Inference"}],"methods":[{"method_slug":"dropout","method_name":"Dropout"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1703.02914","atlas_url":"https://app.syntology.ai/?focus=1703.02914","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}