ConDiSim: Conditional Diffusion Models for Simulation-Based Inference

Mayank Nautiyal, Andreas Hellander, Prashant Singh
Proceedings of The 29th International Conference on Artificial Intelligence and Statistics, PMLR 300:3952-3960, 2026.

Abstract

We present ConDiSim, a conditional diffusion model for simulation-based inference in complex systems with intractable likelihoods. ConDiSim leverages denoising diffusion probabilistic models to approximate posterior distributions, consisting of a forward process that adds Gaussian noise to parameters, and a reverse process learning to denoise, conditioned on observed data. This approach effectively captures complex dependencies and multi-modalities within posteriors. ConDiSim is evaluated across ten benchmark problems and two real-world test problems, where it demonstrates effective posterior approximation accuracy while maintaining computational efficiency and stability in model training. ConDiSim provides a robust and extensible framework for simulation-based inference, well suited to parameter estimation tasks that demand fast methods for handling noisy, time series observations.

Cite this Paper

BibTeX
@InProceedings{pmlr-v300-nautiyal26a, title = { ConDiSim: Conditional Diffusion Models for Simulation-Based Inference }, author = {Nautiyal, Mayank and Hellander, Andreas and Singh, Prashant}, booktitle = {Proceedings of The 29th International Conference on Artificial Intelligence and Statistics}, pages = {3952--3960}, year = {2026}, editor = {Khan, Emtiyaz and Li, Yingzhen and Solin, Arno and Ramdas, Aaditya}, volume = {300}, series = {Proceedings of Machine Learning Research}, month = {02--05 May}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v300/main/assets/nautiyal26a/nautiyal26a.pdf}, url = {https://proceedings.mlr.press/v300/nautiyal26a.html}, abstract = { We present ConDiSim, a conditional diffusion model for simulation-based inference in complex systems with intractable likelihoods. ConDiSim leverages denoising diffusion probabilistic models to approximate posterior distributions, consisting of a forward process that adds Gaussian noise to parameters, and a reverse process learning to denoise, conditioned on observed data. This approach effectively captures complex dependencies and multi-modalities within posteriors. ConDiSim is evaluated across ten benchmark problems and two real-world test problems, where it demonstrates effective posterior approximation accuracy while maintaining computational efficiency and stability in model training. ConDiSim provides a robust and extensible framework for simulation-based inference, well suited to parameter estimation tasks that demand fast methods for handling noisy, time series observations. } }
Endnote
%0 Conference Paper %T ConDiSim: Conditional Diffusion Models for Simulation-Based Inference %A Mayank Nautiyal %A Andreas Hellander %A Prashant Singh %B Proceedings of The 29th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2026 %E Emtiyaz Khan %E Yingzhen Li %E Arno Solin %E Aaditya Ramdas %F pmlr-v300-nautiyal26a %I PMLR %P 3952--3960 %U https://proceedings.mlr.press/v300/nautiyal26a.html %V 300 %X We present ConDiSim, a conditional diffusion model for simulation-based inference in complex systems with intractable likelihoods. ConDiSim leverages denoising diffusion probabilistic models to approximate posterior distributions, consisting of a forward process that adds Gaussian noise to parameters, and a reverse process learning to denoise, conditioned on observed data. This approach effectively captures complex dependencies and multi-modalities within posteriors. ConDiSim is evaluated across ten benchmark problems and two real-world test problems, where it demonstrates effective posterior approximation accuracy while maintaining computational efficiency and stability in model training. ConDiSim provides a robust and extensible framework for simulation-based inference, well suited to parameter estimation tasks that demand fast methods for handling noisy, time series observations.
APA
Nautiyal, M., Hellander, A. & Singh, P.. (2026). ConDiSim: Conditional Diffusion Models for Simulation-Based Inference . Proceedings of The 29th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 300:3952-3960 Available from https://proceedings.mlr.press/v300/nautiyal26a.html.

Related Material