Progressive Tempering Sampler with Diffusion

Severi Rissanen, Ruikang Ouyang, Jiajun He, Wenlin Chen, Markus Heinonen, Arno Solin, José Miguel Hernández-Lobato
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:51724-51746, 2025.

Abstract

Recent research has focused on designing neural samplers that amortize the process of sampling from unnormalized densities. However, despite significant advancements, they still fall short of the state-of-the-art MCMC approach, Parallel Tempering (PT), when it comes to the efficiency of target evaluations. On the other hand, unlike a well-trained neural sampler, PT yields only dependent samples and needs to be rerun—at considerable computational cost—whenever new samples are required. To address these weaknesses, we propose the Progressive Tempering Sampler with Diffusion (PTSD), which trains diffusion models sequentially across temperatures, leveraging the advantages of PT to improve the training of neural samplers. We also introduce a novel method to combine high-temperature diffusion models to generate approximate lower-temperature samples, which are minimally refined using MCMC and used to train the next diffusion model. PTSD enables efficient reuse of sample information across temperature levels while generating well-mixed, uncorrelated samples. Our method significantly improves target evaluation efficiency, outperforming diffusion-based neural samplers.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-rissanen25a, title = {Progressive Tempering Sampler with Diffusion}, author = {Rissanen, Severi and Ouyang, Ruikang and He, Jiajun and Chen, Wenlin and Heinonen, Markus and Solin, Arno and Hern\'{a}ndez-Lobato, Jos\'{e} Miguel}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {51724--51746}, year = {2025}, editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry}, volume = {267}, series = {Proceedings of Machine Learning Research}, month = {13--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/rissanen25a/rissanen25a.pdf}, url = {https://proceedings.mlr.press/v267/rissanen25a.html}, abstract = {Recent research has focused on designing neural samplers that amortize the process of sampling from unnormalized densities. However, despite significant advancements, they still fall short of the state-of-the-art MCMC approach, Parallel Tempering (PT), when it comes to the efficiency of target evaluations. On the other hand, unlike a well-trained neural sampler, PT yields only dependent samples and needs to be rerun—at considerable computational cost—whenever new samples are required. To address these weaknesses, we propose the Progressive Tempering Sampler with Diffusion (PTSD), which trains diffusion models sequentially across temperatures, leveraging the advantages of PT to improve the training of neural samplers. We also introduce a novel method to combine high-temperature diffusion models to generate approximate lower-temperature samples, which are minimally refined using MCMC and used to train the next diffusion model. PTSD enables efficient reuse of sample information across temperature levels while generating well-mixed, uncorrelated samples. Our method significantly improves target evaluation efficiency, outperforming diffusion-based neural samplers.} }
Endnote
%0 Conference Paper %T Progressive Tempering Sampler with Diffusion %A Severi Rissanen %A Ruikang Ouyang %A Jiajun He %A Wenlin Chen %A Markus Heinonen %A Arno Solin %A José Miguel Hernández-Lobato %B Proceedings of the 42nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Aarti Singh %E Maryam Fazel %E Daniel Hsu %E Simon Lacoste-Julien %E Felix Berkenkamp %E Tegan Maharaj %E Kiri Wagstaff %E Jerry Zhu %F pmlr-v267-rissanen25a %I PMLR %P 51724--51746 %U https://proceedings.mlr.press/v267/rissanen25a.html %V 267 %X Recent research has focused on designing neural samplers that amortize the process of sampling from unnormalized densities. However, despite significant advancements, they still fall short of the state-of-the-art MCMC approach, Parallel Tempering (PT), when it comes to the efficiency of target evaluations. On the other hand, unlike a well-trained neural sampler, PT yields only dependent samples and needs to be rerun—at considerable computational cost—whenever new samples are required. To address these weaknesses, we propose the Progressive Tempering Sampler with Diffusion (PTSD), which trains diffusion models sequentially across temperatures, leveraging the advantages of PT to improve the training of neural samplers. We also introduce a novel method to combine high-temperature diffusion models to generate approximate lower-temperature samples, which are minimally refined using MCMC and used to train the next diffusion model. PTSD enables efficient reuse of sample information across temperature levels while generating well-mixed, uncorrelated samples. Our method significantly improves target evaluation efficiency, outperforming diffusion-based neural samplers.
APA
Rissanen, S., Ouyang, R., He, J., Chen, W., Heinonen, M., Solin, A. & Hernández-Lobato, J.M.. (2025). Progressive Tempering Sampler with Diffusion. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:51724-51746 Available from https://proceedings.mlr.press/v267/rissanen25a.html.

Related Material