Diffusion Differentiable Resampling

Jennifer R. Andersson, Zheng Zhao
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:2755-2786, 2026.

Abstract

This paper is concerned with differentiable resampling in the context of sequential Monte Carlo (e.g., particle filtering). Drawing on reparametrisation, we propose a new resampling method that is informative and instantly differentiable, based on a training-free diffusion model surrogate. We theoretically prove that our diffusion resampling method provides a consistent resampling distribution, and we show empirically that it outperforms the state-of-the-art differentiable resampling methods on multiple filtering and parameter estimation benchmarks. Finally, we show that it achieves competitive end-to-end performance when used in learning a complex dynamics-decoder model with high-dimensional image observations.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-andersson26b, title = {Diffusion Differentiable Resampling}, author = {Andersson, Jennifer R. and Zhao, Zheng}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {2755--2786}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/andersson26b/andersson26b.pdf}, url = {https://proceedings.mlr.press/v306/andersson26b.html}, abstract = {This paper is concerned with differentiable resampling in the context of sequential Monte Carlo (e.g., particle filtering). Drawing on reparametrisation, we propose a new resampling method that is informative and instantly differentiable, based on a training-free diffusion model surrogate. We theoretically prove that our diffusion resampling method provides a consistent resampling distribution, and we show empirically that it outperforms the state-of-the-art differentiable resampling methods on multiple filtering and parameter estimation benchmarks. Finally, we show that it achieves competitive end-to-end performance when used in learning a complex dynamics-decoder model with high-dimensional image observations.} }
Endnote
%0 Conference Paper %T Diffusion Differentiable Resampling %A Jennifer R. Andersson %A Zheng Zhao %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-andersson26b %I PMLR %P 2755--2786 %U https://proceedings.mlr.press/v306/andersson26b.html %V 306 %X This paper is concerned with differentiable resampling in the context of sequential Monte Carlo (e.g., particle filtering). Drawing on reparametrisation, we propose a new resampling method that is informative and instantly differentiable, based on a training-free diffusion model surrogate. We theoretically prove that our diffusion resampling method provides a consistent resampling distribution, and we show empirically that it outperforms the state-of-the-art differentiable resampling methods on multiple filtering and parameter estimation benchmarks. Finally, we show that it achieves competitive end-to-end performance when used in learning a complex dynamics-decoder model with high-dimensional image observations.
APA
Andersson, J.R. & Zhao, Z.. (2026). Diffusion Differentiable Resampling. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:2755-2786 Available from https://proceedings.mlr.press/v306/andersson26b.html.

Related Material