Prototype Fusion: A Training-Free Multi-Layer Approach to OOD Detection

• URL: http://arxiv.org/abs/2603.23677v1
• Date: Tue, 24 Mar 2026 19:32:13 GMT
• Title: Prototype Fusion: A Training-Free Multi-Layer Approach to OOD Detection
• Abstract summary: We show that intermediate layers encode equally rich and discriminative information for OOD detection.<n>We propose a model-agnostic approach that leverages internal representations across multiple layers.<n>Our approach delivers robust, architecture-agnostic performance and strong generalization for image classification.
• Score: 0.8394724589095839
• License: http://creativecommons.org/licenses/by-sa/4.0/
• Abstract: Deep learning models are increasingly deployed in safety-critical applications, where reliable out-of-distribution (OOD) detection is essential to ensure robustness. Existing methods predominantly rely on the penultimate-layer activations of neural networks, assuming they encapsulate the most informative in-distribution (ID) representations. In this work, we revisit this assumption to show that intermediate layers encode equally rich and discriminative information for OOD detection. Based on this observation, we propose a simple yet effective model-agnostic approach that leverages internal representations across multiple layers. Our scheme aggregates features from successive convolutional blocks, computes class-wise mean embeddings, and applies L_2 normalization to form compact ID prototypes capturing class semantics. During inference, cosine similarity between test features and these prototypes serves as an OOD score--ID samples exhibit strong affinity to at least one prototype, whereas OOD samples remain uniformly distant. Extensive experiments on state-of-the-art OOD benchmarks across diverse architectures demonstrate that our approach delivers robust, architecture-agnostic performance and strong generalization for image classification. Notably, it improves AUROC by up to 4.41% and reduces FPR by 13.58%, highlighting multi-layer feature aggregation as a powerful yet underexplored signal for OOD detection, challenging the dominance of penultimate-layer-based methods. Our code is available at: https://github.com/sgchr273/cosine-layers.git.

Related papers

• Unified Multi-Layer Subspace Modeling for Cross-Domain OOD Detection [12.06155749281499]
  PRISM is a model-agnostic post-hoc OOD detection method.<n>It fuses intermediate and deep features into a single hierarchical embedding.<n>It captures both in-subspace semantic deviations and off-subspace anomalies.
  arXiv  Detail & Related papers  (2026-09-06T19:00:54Z)
• Sparse Autoencoders for Interpretable Out-of-Distribution Detection [5.034534322537894]
  This paper introduces a novel approach that leverages sparse autoencoders (SAEs) to learn interpretable features from intermediate activations.<n>We propose a new OOD score derived from the cosine similarity between the sparse feature activations of a test sample and the mean activations of ID classes.
  arXiv  Detail & Related papers  (2026-07-13T19:13:40Z)
• Revisiting Logit Distributions for Reliable Out-of-Distribution Detection [73.9121001113687]
  Out-of-distribution (OOD) detection is critical for ensuring the reliability of deep learning models in open-world applications.<n>LogitGap is a novel post-hoc OOD detection method that exploits the relationship between the maximum logit and the remaining logits.<n>We show that LogitGap consistently achieves state-of-the-art performance across diverse OOD detection scenarios and benchmarks.
  arXiv  Detail & Related papers  (2025-10-23T02:16:45Z)
• Mysteries of the Deep: Role of Intermediate Representations in Out of Distribution Detection [0.0]
  Methods treat large pre-trained models as monolithic encoders and rely solely on their final-layer representations for detection.<n>We reveal the textitintermediate layers of pre-trained models, shaped by residual connections that subtly transform input projections.<n>We show that selectively incorporating these intermediate representations can increase the accuracy of OOD detection by up to textbf$10%$ in far-OOD and over textbf$7%$ in near-OOD benchmarks.
  arXiv  Detail & Related papers  (2025-10-07T10:55:47Z)
• Rethinking the Evaluation of Out-of-Distribution Detection: A Sorites Paradox [70.57120710151105]
  Most existing out-of-distribution (OOD) detection benchmarks classify samples with novel labels as the OOD data. Some marginal OOD samples actually have close semantic contents to the in-distribution (ID) sample, which makes determining the OOD sample a Sorites Paradox. We construct a benchmark named Incremental Shift OOD (IS-OOD) to address the issue.
  arXiv  Detail & Related papers  (2024-06-14T09:27:56Z)
• Optimizing OOD Detection in Molecular Graphs: A Novel Approach with Diffusion Models [71.39421638547164]
  We propose to detect OOD molecules by adopting an auxiliary diffusion model-based framework, which compares similarities between input molecules and reconstructed graphs. Due to the generative bias towards reconstructing ID training samples, the similarity scores of OOD molecules will be much lower to facilitate detection. Our research pioneers an approach of Prototypical Graph Reconstruction for Molecular OOD Detection, dubbed as PGR-MOOD and hinges on three innovations.
  arXiv  Detail & Related papers  (2024-04-24T03:25:53Z)
• GROOD: GRadient-Aware Out-of-Distribution Detection [11.511906612904255]
  Out-of-distribution (OOD) detection is crucial for ensuring the reliability of deep learning models in real-world applications.<n>We propose GRadient-aware Out-Of-Distribution detection (GROOD), a method that derives an OOD prototype from synthetic samples and computes class prototypes directly from In-distribution (ID) training data.
  arXiv  Detail & Related papers  (2023-12-22T04:28:43Z)
• Classifier-head Informed Feature Masking and Prototype-based Logit Smoothing for Out-of-Distribution Detection [27.062465089674763]
  Out-of-distribution (OOD) detection is essential when deploying neural networks in the real world. One main challenge is that neural networks often make overconfident predictions on OOD data. We propose an effective post-hoc OOD detection method based on a new feature masking strategy and a novel logit smoothing strategy.
  arXiv  Detail & Related papers  (2023-10-27T12:42:17Z)
• From Global to Local: Multi-scale Out-of-distribution Detection [129.37607313927458]
  Out-of-distribution (OOD) detection aims to detect "unknown" data whose labels have not been seen during the in-distribution (ID) training process. Recent progress in representation learning gives rise to distance-based OOD detection. We propose Multi-scale OOD DEtection (MODE), a first framework leveraging both global visual information and local region details.
  arXiv  Detail & Related papers  (2023-08-20T11:56:25Z)
• General-Purpose Multi-Modal OOD Detection Framework [5.287829685181842]
  Out-of-distribution (OOD) detection identifies test samples that differ from the training data, which is critical to ensuring the safety and reliability of machine learning (ML) systems. We propose a general-purpose weakly-supervised OOD detection framework, called WOOD, that combines a binary classifier and a contrastive learning component. We evaluate the proposed WOOD model on multiple real-world datasets, and the experimental results demonstrate that the WOOD model outperforms the state-of-the-art methods for multi-modal OOD detection.
  arXiv  Detail & Related papers  (2023-07-24T18:50:49Z)
• Improving Out-of-Distribution Detection with Disentangled Foreground and Background Features [23.266183020469065]
  We propose a novel framework that disentangles foreground and background features from ID training samples via a dense prediction approach. It is a generic framework that allows for a seamless combination with various existing OOD detection methods.
  arXiv  Detail & Related papers  (2023-03-15T16:12:14Z)
• How to Exploit Hyperspherical Embeddings for Out-of-Distribution Detection? [22.519572587827213]
  CIDER is a representation learning framework that exploits hyperspherical embeddings for OOD detection. CIDER establishes superior performance, outperforming the latest rival by 19.36% in FPR95.
  arXiv  Detail & Related papers  (2022-03-08T23:44:01Z)
• Triggering Failures: Out-Of-Distribution detection by learning from local adversarial attacks in Semantic Segmentation [76.2621758731288]
  We tackle the detection of out-of-distribution (OOD) objects in semantic segmentation. Our main contribution is a new OOD detection architecture called ObsNet associated with a dedicated training scheme based on Local Adversarial Attacks (LAA) We show it obtains top performances both in speed and accuracy when compared to ten recent methods of the literature on three different datasets.
  arXiv  Detail & Related papers  (2021-08-03T17:09:56Z)
• Enhancing the Generalization for Intent Classification and Out-of-Domain Detection in SLU [70.44344060176952]
  Intent classification is a major task in spoken language understanding (SLU) Recent works have shown that using extra data and labels can improve the OOD detection performance. This paper proposes to train a model with only IND data while supporting both IND intent classification and OOD detection.
  arXiv  Detail & Related papers  (2021-06-28T08:27:38Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.