{"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/arxiv-2605-28589","title":"Thinned Mean Field Langevin Dynamics","arxiv_id":"2605.28589","date":"2026-05-27","proceeding":"ICML","authors":["Zonghao Chen","Heishiro Kanagawa","François-Xavier Briol","Chris J. Oates","Lester Mackey"],"abstract":"Several important learning tasks can be formulated as minimizing an entropy-regularized objective over an appropriate space of probability distributions. Mean-field Langevin dynamics (MFLD) facilitate computation in this general context, casting the minimizer as the invariant distribution of a McKean--Vlasov process, which can be numerically discretized using $N$ particles and thus simulated. However, simulating this interacting particle system has computational complexity of order $N^2$. Motivated by recent research into \\emph{kernel thinning}, we propose \\texttt{KT-MFLD}, in which each particle interacts only with a thinned particle coreset of size $\\mathcal{O}(N^{\\frac{1}{2}})$. \\texttt{KT-MFLD} thus reduces the computational complexity to order $N^{\\frac{3}{2}}$ while, under mild regularity conditions, achieving the same convergence guarantees (up to logarithmic factors) as MFLD. Our theoretical analysis is empirically confirmed on tasks including the training of student-teacher neural networks, quantization with maximum mean discrepancy, and computation of predictively-oriented posteriors in a post-Bayesian framework.","url_abs":"https://arxiv.org/abs/2605.28589","url_pdf":"https://arxiv.org/pdf/2605.28589","source":{"archive":null,"snapshot":"2025-07-28","note":"not in the Papers with Code archive (frozen at the snapshot)","row_kind":"graph","title_abstract_authors_date":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)"},"code_links":[],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2605.28589","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2605.28589"}},"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":[{"mentioned_in_github":null,"is_official":null,"provenance":"deterministic:regex_extraction","mentioned_in_paper":null,"url":"https://github.com/hudsonchen/thinned_mfld","reach":null}],"summary":{"ran":10,"unverified":7},"by_repo_kind":{"found_in_text":{"samples":17,"ran":10,"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":17,"samples":[{"code_sha256_prefix":"b8179bb351b7af3d","entry":"compress","repo":"hudsonchen/thinned_mfld","repo_kind":"found_in_text","path":"utils/kt.py","file_url":"https://github.com/hudsonchen/thinned_mfld/blob/HEAD/utils/kt.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"b8179bb351b7af3d"}},{"code_sha256_prefix":"b768cf9bfb263c52","entry":"create_dir","repo":"hudsonchen/thinned_mfld","repo_kind":"found_in_text","path":"run_mfld.py","file_url":"https://github.com/hudsonchen/thinned_mfld/blob/HEAD/run_mfld.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"b768cf9bfb263c52"}},{"code_sha256_prefix":"5b58f9b860674532","entry":"create_dir","repo":"hudsonchen/thinned_mfld","repo_kind":"found_in_text","path":"run_mfg.py","file_url":"https://github.com/hudsonchen/thinned_mfld/blob/HEAD/run_mfg.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"5b58f9b860674532"}},{"code_sha256_prefix":"6228dc05163b090f","entry":"glorot_normal","repo":"hudsonchen/thinned_mfld","repo_kind":"found_in_text","path":"mfld.py","file_url":"https://github.com/hudsonchen/thinned_mfld/blob/HEAD/mfld.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"6228dc05163b090f"}},{"code_sha256_prefix":"e21f350ed6711088","entry":"initialize","repo":"hudsonchen/thinned_mfld","repo_kind":"found_in_text","path":"mfld.py","file_url":"https://github.com/hudsonchen/thinned_mfld/blob/HEAD/mfld.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"e21f350ed6711088"}},{"code_sha256_prefix":"bc3056a4746e2cef","entry":"kme_RBF_Gaussian_func","repo":"hudsonchen/thinned_mfld","repo_kind":"found_in_text","path":"utils/kernel.py","file_url":"https://github.com/hudsonchen/thinned_mfld/blob/HEAD/utils/kernel.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"bc3056a4746e2cef"}},{"code_sha256_prefix":"c76051bf80f81aa0","entry":"kme_double_RBF_Gaussian","repo":"hudsonchen/thinned_mfld","repo_kind":"found_in_text","path":"utils/kernel.py","file_url":"https://github.com/hudsonchen/thinned_mfld/blob/HEAD/utils/kernel.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"c76051bf80f81aa0"}},{"code_sha256_prefix":"1c8333152ea5c4c8","entry":"lvfunc","repo":"hudsonchen/thinned_mfld","repo_kind":"found_in_text","path":"utils/lotka_volterra.py","file_url":"https://github.com/hudsonchen/thinned_mfld/blob/HEAD/utils/lotka_volterra.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"1c8333152ea5c4c8"}},{"code_sha256_prefix":"cff708f580a95837","entry":"sample_rho0","repo":"hudsonchen/thinned_mfld","repo_kind":"found_in_text","path":"run_mfg.py","file_url":"https://github.com/hudsonchen/thinned_mfld/blob/HEAD/run_mfg.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"cff708f580a95837"}},{"code_sha256_prefix":"acad404b81855d83","entry":"terminal_grad_psi","repo":"hudsonchen/thinned_mfld","repo_kind":"found_in_text","path":"run_mfg.py","file_url":"https://github.com/hudsonchen/thinned_mfld/blob/HEAD/run_mfg.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"acad404b81855d83"}},{"code_sha256_prefix":"cbb4621dcf5e413c","entry":"eval_mmd_flow","repo":"hudsonchen/thinned_mfld","repo_kind":"found_in_text","path":"utils/evaluate.py","file_url":"https://github.com/hudsonchen/thinned_mfld/blob/HEAD/utils/evaluate.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"cbb4621dcf5e413c"}},{"code_sha256_prefix":"35a125973e0bc629","entry":"eval_vlm","repo":"hudsonchen/thinned_mfld","repo_kind":"found_in_text","path":"utils/evaluate.py","file_url":"https://github.com/hudsonchen/thinned_mfld/blob/HEAD/utils/evaluate.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"35a125973e0bc629"}},{"code_sha256_prefix":"7437e7146b7c6897","entry":"load_boston","repo":"hudsonchen/thinned_mfld","repo_kind":"found_in_text","path":"utils/datasets.py","file_url":"https://github.com/hudsonchen/thinned_mfld/blob/HEAD/utils/datasets.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"7437e7146b7c6897"}},{"code_sha256_prefix":"cdd43ea56e03f4f5","entry":"load_covertype","repo":"hudsonchen/thinned_mfld","repo_kind":"found_in_text","path":"utils/datasets.py","file_url":"https://github.com/hudsonchen/thinned_mfld/blob/HEAD/utils/datasets.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"cdd43ea56e03f4f5"}},{"code_sha256_prefix":"73566abcfe6ceb54","entry":"make_teacher_fn","repo":"hudsonchen/thinned_mfld","repo_kind":"found_in_text","path":"utils/datasets.py","file_url":"https://github.com/hudsonchen/thinned_mfld/blob/HEAD/utils/datasets.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"73566abcfe6ceb54"}},{"code_sha256_prefix":"aa946eba0a62c4fd","entry":"save_animation_2d","repo":"hudsonchen/thinned_mfld","repo_kind":"found_in_text","path":"utils/evaluate.py","file_url":"https://github.com/hudsonchen/thinned_mfld/blob/HEAD/utils/evaluate.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"aa946eba0a62c4fd"}},{"code_sha256_prefix":"9a76f184656d66ca","entry":"uncentered_matern_32_kernel","repo":"hudsonchen/thinned_mfld","repo_kind":"found_in_text","path":"mfld.py","file_url":"https://github.com/hudsonchen/thinned_mfld/blob/HEAD/mfld.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"9a76f184656d66ca"}}]},"arxiv_metadata":{"licence":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)","fields":["title","abstract","authors","date"],"primary_category":"cs.LG","source":"arxiv_2026.jsonl"},"syntology_extracted_results":null}