HSSM is a Python toolbox for hierarchical Bayesian modeling of choice and response-time data with sequential sampling models. It supports trial-wise and hierarchical regression, reinforcement-learning models, posterior diagnostics, model comparison, and custom likelihoods through a high-level PyMC and Bambi interface. HSSM is a BRAINSTORM project at Brown University.
Use Python 3.12, 3.13, or 3.14 in a fresh environment:
pip install hssmThe installation guide covers uv, CUDA extras, Colab, development installs, and troubleshooting.
The HSSM documentation is the canonical source for durable guidance. Begin with the quickstart, then follow the main tutorial. The ecosystem map explains when work belongs in HSSM or one of its sibling projects.
- Read the contribution guide and local development setup.
- Ask modeling questions in GitHub Discussions.
- Report bugs and request features through GitHub Issues.
Please cite Fengler et al., HSSM: A Widely Applicable Toolbox for Hierarchical Bayesian Neurocognitive Modeling (paper DOI). For version-specific software citation, use the Zenodo archive.
HSSM carries the Brown University license in LICENSE. Copyright 2023 Brown University. All Rights Reserved.