sbi reloaded: a toolkit for simulation-based inference workflows

Scientists and engineers use simulators to model empirically observed phenomena.However, tuning the parameters of a simulator to ensure its outputs match observed data presents a significant challenge.Simulation-based inference (SBI) addresses this by enabling Bayesian inference for simulators, identifying parameters that match observed data and align with prior knowledge.Unlike traditional Bayesian inference, SBI only needs access to simulations from the model and does not require evaluations of the likelihood-function.In addition, SBI algorithms do not require gradients through the simulator, allow for massive parallelization of simulations, and can perform inference for different observations without further simulations or training, thereby amortizing inference.Over the past years, we have developed, maintained, and extended sbi, a PyTorch-based package that implements Bayesian SBI algorithms based on neural networks.The sbi toolkit implements a wide range of inference methods, neural network architectures, sampling methods, and diagnostic tools.In addition, it provides well-tested default settings but also offers flexibility to fully customize every step of the simulation-based inference workflow.Taken together, the sbi toolkit enables scientists and engineers to apply state-of-the-art SBI methods to black-box simulators, opening up new possibilities for aligning simulations with empirically observed data.

Authors

Institutions

Publication Details

Journal
The Journal of Open Source Software
Published
2025-04-08
DOI
https://doi.org/10.21105/joss.07754
Citations
11
Primary Topic
Scientific Computing and Data Management
Type
article
Field-Weighted Citation Impact
45.86

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

sbi reloaded: a toolkit for simulation-based inference workflows

Jonas Beck, Cornelius Schröder, Miriam Herold, Abolfazl Ziaeemehr et al.
11 citations
The Journal of Open Source Software
Scientific Computing and Data Management
45.86
article

sbi reloaded: a toolkit for simulation-based inference workflows

Jonas Beck, Cornelius Schröder, Miriam Herold, Abolfazl Ziaeemehr, B. Miller, Zinovia Stefanidi, Jaivardhan Kapoor, Pedro Luiz Coelho Rodrigues, Julius Vetter, Guy Moss, Auguste Schulz, Jan-Matthis Lueckmann, Bálint Mucsányi, Matthijs Pals, Pedro J. Gonçalves, Jan Boelts, Fabio Muratore, Theo Gruner, Álvaro Tejero-Cantero, Peter Steinbach, Julia Linhart, Thomas Moreau, Janne K. Lappalainen, Conor Durkan, Jakob H. Macke, Felix Pei, Michael Deistler, Manuel Gloeckler, Sebastian Bischoff, David Greenberg, Richard Gao, Nastya Krouglova, Faried Abu Zaid
article en
11 citations

Abstract

Scientists and engineers use simulators to model empirically observed phenomena.However, tuning the parameters of a simulator to ensure its outputs match observed data presents a significant challenge.Simulation-based inference (SBI) addresses this by enabling Bayesian inference for simulators, identifying parameters that match observed data and align with prior knowledge.Unlike traditional Bayesian inference, SBI only needs access to simulations from the model and does not require evaluations of the likelihood-function.In addition, SBI algorithms do not require gradients through the simulator, allow for massive parallelization of simulations, and can perform inference for different observations without further simulations or training, thereby amortizing inference.Over the past years, we have developed, maintained, and extended sbi, a PyTorch-based package that implements Bayesian SBI algorithms based on neural networks.The sbi toolkit implements a wide range of inference methods, neural network architectures, sampling methods, and diagnostic tools.In addition, it provides well-tested default settings but also offers flexibility to fully customize every step of the simulation-based inference workflow.Taken together, the sbi toolkit enables scientists and engineers to apply state-of-the-art SBI methods to black-box simulators, opening up new possibilities for aligning simulations with empirically observed data.

The Journal of Open Source SoftwareVol. 10(108)
Institut polytechnique de Grenoble (FR), BMW (Germany) (DE), Google (United States) (US), Centre National de la Recherche Scientifique (FR), Institut national de recherche en sciences et technologies du numérique (FR), Inserm (FR), Aix-Marseille Université (FR), Commissariat à l'Énergie Atomique et aux Énergies Alternatives (FR), Université Paris-Saclay (FR), Helmholtz-Zentrum Dresden-Rossendorf (DE), Technische Universität Darmstadt (DE), Universität der Bundeswehr München (DE), Neuroscience Institute (IT), Innovation Cluster (Canada) (CA), Centre Inria de Saclay (FR), CEA Paris-Saclay (FR), Bernstein Center for Computational Neuroscience Tübingen (DE), Institute of Marine and Coastal Research (AR), Max Planck Institute for Intelligent Systems (DE), Hertie Institute for Clinical Brain Research (DE), Neuroelectronics Research Flanders (BE), Robert Bosch (Taiwan) (TW), Laboratoire Jean Kuntzmann (FR), BMW Group (Germany) (DE), Tübingen AI Center, Carnegie Mellon University (US), University of Tübingen (DE), University of Amsterdam (NL), Université Grenoble Alpes (FR), University of Edinburgh (GB), KU Leuven (BE)
European Commission, Deutsche Forschungsgemeinschaft, Bundesministerium für Bildung und Forschung, Engineering and Physical Sciences Research Council
Openalex Percentile: Top 0%
Scientific Computing and Data Management
45.86
DOIOpen Access PDFOpenAlex
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.