Density of States Estimation for Out-of-Distribution Detection

TLDR

Probabilistic model likelihoods often lack specificity for out‑of‑distribution detection, sometimes assigning higher likelihoods to OOD data than to in‑distribution data. The authors propose DoSE, a density of states estimator, to address this limitation. DoSE computes the frequency of model statistics using nonparametric density estimators such as KDE and one‑class SVM, thereby flagging test points with low typicality as anomalous. The method achieves state‑of‑the‑art unsupervised OOD detection without requiring labeled data or OOD examples and can be applied modularly to any trained model.

Abstract

Perhaps surprisingly, recent studies have shown probabilistic model likelihoods have poor specificity for out-of-distribution (OOD) detection and often assign higher likelihoods to OOD data than in-distribution data. To ameliorate this issue we propose DoSE, the density of states estimator. Drawing on the statistical physics notion of ``density of states,'' the DoSE decision rule avoids direct comparison of model probabilities, and instead utilizes the ``probability of the model probability,'' or indeed the frequency of any reasonable statistic. The frequency is calculated using nonparametric density estimators (e.g., KDE and one-class SVM) which measure the typicality of various model statistics given the training data and from which we can flag test points with low typicality as anomalous. Unlike many other methods, DoSE requires neither labeled data nor OOD examples. DoSE is modular and can be trivially applied to any existing, trained model. We demonstrate DoSE's state-of-the-art performance against other unsupervised OOD detectors on previously established ``hard'' benchmarks.