Search papers, labs, and topics across Lattice.
This paper introduces the Hierarchical Likelihood Vector (HLV) and the Hierarchical Likelihood Distance (HLD) as novel methods for out-of-distribution (OOD) detection using Probabilistic Circuits (PCs). By leveraging the hierarchical structure of PCs, the authors show that HLD serves as an integral probability metric that enhances OOD detection performance without requiring held-out in-distribution data at deployment. Experimental results on tabular and MNIST datasets reveal significant improvements over existing OOD detection methods, highlighting the effectiveness of utilizing hierarchical probabilistic summaries for better localization of distribution shifts.
Hierarchical probabilistic summaries can dramatically enhance out-of-distribution detection, outperforming traditional methods without needing extra in-distribution data.
Probabilistic Circuits (PCs) are tractable generative models whose internal nodes encode a hierarchy of probabilistic sum- maries over different variable scopes. Existing PC-based out- of-distribution (OOD) detection methods ignore this hierar- chy, reducing the entire circuit to the scalar likelihood (or its uncertainty) computed at the root. We introduce Hierar- chical Likelihood Vector (HLV), a representation whose en- tries are the likelihoods associated with selected PC nodes and define the Hierarchical Likelihood Distance (HLD), a PC-induced pseudo-metric that compares the probability dis- tributions through the expectations of their HLVs. We show that HLD is an integral probability metric over a function class naturally induced by the PC and develop a principled goodness-of-fit hypothesis test for unsupervised OOD detec- tion. Unlike existing approaches, the trained PC alone serves as the representation of the in-distribution: no held-out in- distribution data are required at deployment. We further show that the quantities required by the hypothesis test can be com- puted exactly, directly from the trained circuit, yielding an ap- proximate analytic decision threshold. Experiments on tabular and MNIST datasets demonstrate that exploiting the hierarchi- cal probabilistic summaries encoded through the PC improve OOD detection over root-likelihood, uncertainty-, typicality- and kernel-based baselines, while naturally localizing distri- bution shifts to the PC nodes responsible for the shift.
