Scalable Simulation-Based Model Inference with Test-Time Complexity Control

Manuel Gloeckler, J.P. Manzano-Patrón, Stamatios Sotiropoulos, Cornelius Schröder, Jakob H. Macke
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:35227-35268, 2026.

Abstract

Simulation plays a central role in scientific discovery. In many applications, the bottleneck is no longer running a simulator—it is choosing among large families of plausible simulators, each corresponding to different forward models/hypotheses consistent with observations. Over large model families, classical Bayesian workflows for model-selection are impractical. Furthermore, amortized model-selection methods typically hard-code a fixed model prior—or complexity penalty—at training time, requiring users to commit to a particular parsimony assumption before seeing the data. We introduce PRISM, a simulation-based encoder-decoder that infers a joint posterior over both discrete model structures and associated continuous parameters, while enabling test-time control of model complexity via a tunable model prior that the network is conditioned on. We show that PRISM scales to families with combinatorially many (up to billions of) model instantiations on a synthetic symbolic regression task. As a scientific application, we evaluate PRISM on biophysical modeling for diffusion MRI data, showing the ability to perform model selection across several multi-compartment models, on both synthetic and in-vivo neuroimaging data.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-gloeckler26a, title = {Scalable Simulation-Based Model Inference with Test-Time Complexity Control}, author = {Gloeckler, Manuel and Manzano-Patr\'{o}n, J.P. and Sotiropoulos, Stamatios and Schr\"{o}der, Cornelius and Macke, Jakob H.}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {35227--35268}, 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/gloeckler26a/gloeckler26a.pdf}, url = {https://proceedings.mlr.press/v306/gloeckler26a.html}, abstract = {Simulation plays a central role in scientific discovery. In many applications, the bottleneck is no longer running a simulator—it is choosing among large families of plausible simulators, each corresponding to different forward models/hypotheses consistent with observations. Over large model families, classical Bayesian workflows for model-selection are impractical. Furthermore, amortized model-selection methods typically hard-code a fixed model prior—or complexity penalty—at training time, requiring users to commit to a particular parsimony assumption before seeing the data. We introduce PRISM, a simulation-based encoder-decoder that infers a joint posterior over both discrete model structures and associated continuous parameters, while enabling test-time control of model complexity via a tunable model prior that the network is conditioned on. We show that PRISM scales to families with combinatorially many (up to billions of) model instantiations on a synthetic symbolic regression task. As a scientific application, we evaluate PRISM on biophysical modeling for diffusion MRI data, showing the ability to perform model selection across several multi-compartment models, on both synthetic and in-vivo neuroimaging data.} }
Endnote
%0 Conference Paper %T Scalable Simulation-Based Model Inference with Test-Time Complexity Control %A Manuel Gloeckler %A J.P. Manzano-Patrón %A Stamatios Sotiropoulos %A Cornelius Schröder %A Jakob H. Macke %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-gloeckler26a %I PMLR %P 35227--35268 %U https://proceedings.mlr.press/v306/gloeckler26a.html %V 306 %X Simulation plays a central role in scientific discovery. In many applications, the bottleneck is no longer running a simulator—it is choosing among large families of plausible simulators, each corresponding to different forward models/hypotheses consistent with observations. Over large model families, classical Bayesian workflows for model-selection are impractical. Furthermore, amortized model-selection methods typically hard-code a fixed model prior—or complexity penalty—at training time, requiring users to commit to a particular parsimony assumption before seeing the data. We introduce PRISM, a simulation-based encoder-decoder that infers a joint posterior over both discrete model structures and associated continuous parameters, while enabling test-time control of model complexity via a tunable model prior that the network is conditioned on. We show that PRISM scales to families with combinatorially many (up to billions of) model instantiations on a synthetic symbolic regression task. As a scientific application, we evaluate PRISM on biophysical modeling for diffusion MRI data, showing the ability to perform model selection across several multi-compartment models, on both synthetic and in-vivo neuroimaging data.
APA
Gloeckler, M., Manzano-Patrón, J., Sotiropoulos, S., Schröder, C. & Macke, J.H.. (2026). Scalable Simulation-Based Model Inference with Test-Time Complexity Control. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:35227-35268 Available from https://proceedings.mlr.press/v306/gloeckler26a.html.

Related Material