Our paper “Beyond Binary Out-of-Distribution Detection: Characterizing Distributional Shifts with Multi-Statistic Diffusion Trajectories” has been accepted for publication at the 29th International Conference on Artificial Intelligence and Statistics: AISTATS 2026. It will be presented at the conference in early May in Tangier, Morocco.
In the paper, we ask the question whether out-of-distribution detection can go beyond a typically formulated binary question. That is, rather than wanting to only know whether a new datapoint is out-of-distribution, we would also like to be able to distinguish different kinds of detected shift to separate potential sources. We achieve this separation by introducing a diffusion based statistical characterization framework. In short, our framework can distinguish different kinds of out-of-distribution data by leveraging the time-signal of the diffusion process, allowing decisions to be based on unique temporal signatures rather than mere decisions based on “in” vs. “out”.
For more information, read the full paper. The abstract is provided below:
Detecting out-of-distribution (OOD) data is critical for machine learning, be it for safety reasons or to enable open-ended learning. However, beyond mere detection, choosing an appropriate course of action typically hinges on the type of OOD data encountered. Unfortunately, the latter is generally not distinguished in practice, as modern OOD detection methods collapse distributional shifts into single scalar outlier scores. This work argues that scalar-based methods are thus insufficient for OOD data to be properly contextualized and prospectively exploited, a limitation we overcome with the introduction of DISC: Diffusion-based Statistical Characterization. DISC leverages the iterative denoising process of diffusion models to extract a rich, multi-dimensional feature vector that captures statistical discrepancies across multiple noise levels. Extensive experiments on image and tabular benchmarks show that DISC matches or surpasses state-of-the-art detectors for OOD detection and, crucially, also classifies OOD type, a capability largely absent from prior work. As such, our work enables a shift from simple binary OOD detection to a more granular detection.