{"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/boosting-black-box-variational-inference","title":"Boosting Black Box Variational Inference","arxiv_id":"1806.02185","date":"2018-06-06","proceeding":"NeurIPS 2018 12","authors":["Francesco Locatello","Gideon Dresdner","Rajiv Khanna","Isabel Valera","Gunnar Rätsch"],"abstract":"Approximating a probability density in a tractable manner is a central task\nin Bayesian statistics. Variational Inference (VI) is a popular technique that\nachieves tractability by choosing a relatively simple variational family.\nBorrowing ideas from the classic boosting framework, recent approaches attempt\nto \\emph{boost} VI by replacing the selection of a single density with a\ngreedily constructed mixture of densities. In order to guarantee convergence,\nprevious works impose stringent assumptions that require significant effort for\npractitioners. Specifically, they require a custom implementation of the greedy\nstep (called the LMO) for every probabilistic model with respect to an\nunnatural variational family of truncated distributions. Our work fixes these\nissues with novel theoretical and algorithmic insights. On the theoretical\nside, we show that boosting VI satisfies a relaxed smoothness assumption which\nis sufficient for the convergence of the functional Frank-Wolfe (FW) algorithm.\nFurthermore, we rephrase the LMO problem and propose to maximize the Residual\nELBO (RELBO) which replaces the standard ELBO optimization in VI. These\ntheoretical enhancements allow for black box implementation of the boosting\nsubroutine. Finally, we present a stopping criterion drawn from the duality gap\nin the classic FW analyses and exhaustive experiments to illustrate the\nusefulness of our theoretical and algorithmic contributions.","url_abs":"http://arxiv.org/abs/1806.02185v5","url_pdf":"http://arxiv.org/pdf/1806.02185v5.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":"boosting-black-box-variational-inference","repo_url":"https://github.com/ratschlab/boosting-bbvi","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"variational-inference","task_name":"Variational Inference"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1806.02185","atlas_url":"https://app.syntology.ai/?focus=1806.02185","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1806.02185"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/ratschlab/boosting-bbvi","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"unverified":4},"by_repo_kind":{"official":{"samples":4,"ran":0,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"87818452d653055a","entry":"decay_linear","repo":"ratschlab/boosting-bbvi","repo_kind":"official","path":"boosting_bbvi/core/utils.py","file_url":"https://github.com/ratschlab/boosting-bbvi/blob/HEAD/boosting_bbvi/core/utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"87818452d653055a"}},{"code_sha256_prefix":"9d0a63c5a9ff04f9","entry":"decay_log","repo":"ratschlab/boosting-bbvi","repo_kind":"official","path":"boosting_bbvi/core/utils.py","file_url":"https://github.com/ratschlab/boosting-bbvi/blob/HEAD/boosting_bbvi/core/utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9d0a63c5a9ff04f9"}},{"code_sha256_prefix":"ddb327ccec5988fc","entry":"decay_squared","repo":"ratschlab/boosting-bbvi","repo_kind":"official","path":"boosting_bbvi/core/utils.py","file_url":"https://github.com/ratschlab/boosting-bbvi/blob/HEAD/boosting_bbvi/core/utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"ddb327ccec5988fc"}},{"code_sha256_prefix":"2b9e1b70201df6c3","entry":"parse_filename","repo":"ratschlab/boosting-bbvi","repo_kind":"official","path":"boosting_bbvi/plots/plot_utils.py","file_url":"https://github.com/ratschlab/boosting-bbvi/blob/HEAD/boosting_bbvi/plots/plot_utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"2b9e1b70201df6c3"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}