Tippett-minimum Fusion of Representation-space Diffusion Models for Multi-Encoder Out-of-Distribution Detection
The paper presents a novel approach for out-of-distribution (OOD) detection using a multi-encoder fusion of representation-space diffusion models. The proposed method, EncMin2L, effectively combines and calibrates likelihood detectors without requiring OOD labels, achieving significant performance improvements. It demonstrates superior results across various distribution shifts while maintaining lower parameter costs compared to existing models.
- ▪The study addresses OOD detection across multiple types of distribution shifts.
- ▪EncMin2L combines per-encoder diffusion-based likelihood detectors without OOD labels.
- ▪The method outperforms state-of-the-art OOD detectors with a 2.3 times lower parameter cost.
Opening excerpt (first ~120 words) tap to expand
Computer Science > Machine Learning arXiv:2605.20502 (cs) [Submitted on 19 May 2026] Title:Tippett-minimum Fusion of Representation-space Diffusion Models for Multi-Encoder Out-of-Distribution Detection Authors:Neelkamal Bhuyan View a PDF of the paper titled Tippett-minimum Fusion of Representation-space Diffusion Models for Multi-Encoder Out-of-Distribution Detection, by Neelkamal Bhuyan View PDF HTML (experimental) Abstract:We address out-of-distribution (OOD) detection across the full spectrum of distribution shifts -- global domain changes, semantic divergence, texture differences, and covariate corruptions -- through a multi-encoder fusion of per-encoder representation-space diffusion models (RDMs).
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv cs.AI.
Discussion
0 commentsMore from arXiv cs.AI
