Bde: A Python Package for Bayesian Deep Ensembles via MILE

MCML Authors

Emanuel Sommer

Dr.

→ Group David Rügamer
 Statistics, Data Science and Machine Learning

David Rügamer

Prof. Dr.

Core PI

Statistics, Data Science and Machine Learning

Abstract

bde is a Python package designed to bring state-of-the-art sampling-based Bayesian Deep Learning (BDL) to practitioners and researchers. The package combines the speed and highperformance capabilities of JAX (Bradbury et al., 2018) and BlackJAX (Cabezas et al., 2024) with the user-friendly API of scikit-learn (Pedregosa et al., 2011). It targets tabular supervised tasks, including distributional regression and (multi-class) classification, providing a seamless interface for Bayesian Deep Ensembles (BDEs) (Sommer et al., 2024) specifically implementing Microcanonical Langevin Ensembles (MILE) (Sommer et al., 2025). <br>The workflow of bde implements the two-stage BDE inference process of MILE. First, it optimizes n_members independent instances of a flexibly configurable feed-forward neural network using regularized empirical risk minimization (with the negative log-likelihood as loss) via the AdamW optimizer (Loshchilov & Hutter, 2019). Second, it transitions to a sampling phase using Microcanonical Langevin Monte Carlo (Robnik et al., 2023; Robnik & Seljak, 2024), enhanced with a tuning phase adapted for Bayesian neural networks. This combination is referred to as MILE (Sommer et al., 2025). In essence, the optimization of the ensemble of neural networks first finds diverse high-likelihood modes, from which sampling then explores local posterior structure. This process generates an ensemble of samples (models) that constitute an implicit posterior approximation.

article AAS+26

---

The Journal of Open Source Software

11.124. Aug. 2026.

---

Authors

V. Arvanitis • A. Aslanidis • E. Sommer • D. Rügamer

---

Links

DOI GitHub

---

Research Area

 A1 | Statistical Foundations & Explainability