{"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/scaling-ensemble-distribution-distillation-to","title":"Scaling Ensemble Distribution Distillation to Many Classes with Proxy Targets","arxiv_id":"2105.06987","date":"2021-05-14","proceeding":"NeurIPS 2021 12","authors":["Max Ryabinin","Andrey Malinin","Mark Gales"],"abstract":"Ensembles of machine learning models yield improved system performance as well as robust and interpretable uncertainty estimates; however, their inference costs may often be prohibitively high. \\emph{Ensemble Distribution Distillation} is an approach that allows a single model to efficiently capture both the predictive performance and uncertainty estimates of an ensemble. For classification, this is achieved by training a Dirichlet distribution over the ensemble members' output distributions via the maximum likelihood criterion. Although theoretically principled, this criterion exhibits poor convergence when applied to large-scale tasks where the number of classes is very high. In our work, we analyze this effect and show that the Dirichlet log-likelihood criterion classes with low probability induce larger gradients than high-probability classes. This forces the model to focus on the distribution of the ensemble tail-class probabilities. We propose a new training objective that minimizes the reverse KL-divergence to a \\emph{Proxy-Dirichlet} target derived from the ensemble. This loss resolves the gradient issues of Ensemble Distribution Distillation, as we demonstrate both theoretically and empirically on the ImageNet and WMT17 En-De datasets containing 1000 and 40,000 classes, respectively.","url_abs":"https://arxiv.org/abs/2105.06987v1","url_pdf":"https://arxiv.org/pdf/2105.06987v1.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":"scaling-ensemble-distribution-distillation-to","repo_url":"https://github.com/KaosEngineer/structured-uncertainty","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2105.06987","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2105.06987"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+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/KaosEngineer/structured-uncertainty","reach":null}],"summary":{"ran_honours":1,"ran_draft_wrong":3},"by_repo_kind":{"listed":{"samples":4,"ran":4,"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":4,"samples":[{"code_sha256_prefix":"2b7c186c9b54effb","entry":"get_sentence_wer","repo":"KaosEngineer/structured-uncertainty","repo_kind":"listed","path":"examples/structured_uncertainty/assessment/compute_cross_metrics.py","file_url":"https://github.com/KaosEngineer/structured-uncertainty/blob/HEAD/examples/structured_uncertainty/assessment/compute_cross_metrics.py","link_basis":"first_harvest_node","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"well_formed","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"2b7c186c9b54effb"}},{"code_sha256_prefix":"5b36af4674a8d16d","entry":"load_text","repo":"KaosEngineer/structured-uncertainty","repo_kind":"listed","path":"examples/structured_uncertainty/assessment/compute_cross_metrics.py","file_url":"https://github.com/KaosEngineer/structured-uncertainty/blob/HEAD/examples/structured_uncertainty/assessment/compute_cross_metrics.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"5b36af4674a8d16d"}},{"code_sha256_prefix":"5cb5b04d985b725c","entry":"load_uncertainties","repo":"KaosEngineer/structured-uncertainty","repo_kind":"listed","path":"examples/structured_uncertainty/assessment/ood_detection.py","file_url":"https://github.com/KaosEngineer/structured-uncertainty/blob/HEAD/examples/structured_uncertainty/assessment/ood_detection.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"5cb5b04d985b725c"}},{"code_sha256_prefix":"84515154843f0ccf","entry":"load_uncertainties_calibrate","repo":"KaosEngineer/structured-uncertainty","repo_kind":"listed","path":"examples/structured_uncertainty/assessment/ood_detection.py","file_url":"https://github.com/KaosEngineer/structured-uncertainty/blob/HEAD/examples/structured_uncertainty/assessment/ood_detection.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"84515154843f0ccf"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}