Probabilistic Circuits That Know What They Don't Know

Fabrizio Ventola, Steven Braun, Zhongjie Yu, Martin Mundt, Kristian Kersting
2/13/2023
cs.LGcs.AI

No official code found. Best implementation: ml-research/tractable-dropout-inference (8 stars, last commit 2023-07-27).

Abstract

Probabilistic circuits (PCs) are models that allow exact and tractable probabilistic inference. In contrast to neural networks, they are often assumed to be well-calibrated and robust to out-of-distribution (OOD) data. In this paper, we show that PCs are in fact not robust to OOD data, i.e., they don't know what they don't know. We then show how this challenge can be overcome by model uncertainty quantification. To this end, we propose tractable dropout inference (TDI), an inference procedure to estimate uncertainty by deriving an analytical solution to Monte Carlo dropout (MCD) through variance propagation. Unlike MCD in neural networks, which comes at the cost of multiple network evaluations, TDI provides tractable sampling-free uncertainty estimates in a single forward pass. TDI improves the robustness of PCs to distribution shift and OOD data, demonstrated through a series of experiments evaluating the classification confidence and uncertainty estimates on real-world data.

View on arXivView PDF

Code Implementations(1)

ml-research/tractable-dropout-inferenceCommunity100%

Code for our paper "Probabilistic Circuits That Know What They Don't Know".

83PythonFeb 13, 20233 years agoNOASSERTION

Similar Papers

Unreliable Uncertainty Estimates with Monte Carlo Dropout

88% match2025

Calibrated and uncertain? Evaluating uncertainty estimates in binary classification models

84% match2025

Dropout Injection at Test Time for Post Hoc Uncertainty Quantification in Neural Networks

84% match1 repo102023

Probabilistic Skip Connections for Deterministic Uncertainty Quantification in Deep Neural Networks

84% match2025

Efficient Uncertainty Estimation in Spiking Neural Networks via MC-dropout

83% match2023

Uncertainty Estimation using Variance-Gated Distributions

83% match2025

Cite this paper

@article{ventola2023probabilistic,
  title  = {Probabilistic Circuits That Know What They Don't Know},
  author = {Fabrizio Ventola and Steven Braun and Zhongjie Yu and Martin Mundt and Kristian Kersting},
  year   = {2023},
  eprint = {2302.06544},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2302.06544v3}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2302.06544.svg)](https://codeofpaper.com/paper/2302.06544)

Discussion