Autoencoding Probabilistic Circuits

Authors: Steven Braun, Sahil Sidheekh, Sriraam Natarajan, Antonio Vergari, Martin Mundt, Kristian Kersting

Project URL: https://openreview.net/forum?id=Y7dRmpGiHj

Paper: https://openreview.net/pdf?id=Y7dRmpGiHj

Abstract: Probabilistic Circuits (PCs) enable exact tractable inference, yet their application to representation learning remains underexplored. We introduce Autoencoding Probabilistic Circuits (APCs), a novel framework leveraging PC tractability to explicitly model probabilistic embeddings. APCs jointly model data and latent representations, obtaining embeddings via probabilistic inference using a PC encoder, which is integrated with a neural decoder in an end-to-end trainable hybrid architecture enabled by differentiable sampling. Empirical evaluations demonstrate APCs outperform existing PC-based autoencoding methods in reconstruction quality, generate embeddings competitive with neural autoencoders, and exhibit superior robustness to missing data without requiring imputation. These results establish APCs as a powerful and flexible representation learning method, exploiting PC inference capabilities for robust applications including out-of-distribution detection and knowledge distillation.

@inproceedings{ braun2025autoencoding, title={Autoencoding Probabilistic Circuits}, author={Steven Braun and Sahil Sidheekh and Sriraam Natarajan and Antonio Vergari and Martin Mundt and Kristian Kersting}, booktitle={Eighth Workshop on Tractable Probabilistic Modeling}, year={2025}, url={https://openreview.net/forum?id=zgxp6sNpie} }