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

arXiv: https://arxiv.org/abs/2603.23677 · HTML: https://mlsanity.com/paper/2603.23677/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 20 of its 23 references that are in the map.
- First posted on arXiv 24 Mar 2026.

## Builds on (most cited first, 20 of 20)

- [A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks](https://mlsanity.com/paper/1610.02136/) (2016)
- [Energy-based Out-of-distribution Detection](https://mlsanity.com/paper/2010.03759/) (2020)
- [Out-of-Distribution Detection with Deep Nearest Neighbors](https://mlsanity.com/paper/2204.06507/) (2022)
- [ReAct: Out-of-distribution Detection With Rectified Activations](https://mlsanity.com/paper/2111.12797/) (2021)
- [Scaling Out-of-Distribution Detection for Real-World Settings](https://mlsanity.com/paper/1911.11132/) (2019)
- [On the Importance of Gradients for Detecting Distributional Shifts in the Wild](https://mlsanity.com/paper/2110.00218/) (2021)
- [SSD: A Unified Framework for Self-Supervised Outlier Detection](https://mlsanity.com/paper/2103.12051/) (2021)
- [Nearest Neighbor Guidance for Out-of-Distribution Detection](https://mlsanity.com/paper/2309.14888/) (2023)
- [Neural Mean Discrepancy for Efficient Out-of-Distribution Detection](https://mlsanity.com/paper/2104.11408/) (2021)
- [NECO: NEural Collapse Based Out-of-distribution detection](https://mlsanity.com/paper/2310.06823/) (2023)
- [Detecting Out-of-Distribution Through the Lens of Neural Collapse](https://mlsanity.com/paper/2311.01479/) (2023)
- [Out-of-Distribution Detection by Leveraging Between-Layer Transformation Smoothness](https://mlsanity.com/paper/2310.02832/) (2023)
- [Pursuing Feature Separation based on Neural Collapse for Out-of-Distribution Detection](https://mlsanity.com/paper/2405.17816/) (2024)
- [Controlling Neural Collapse Enhances Out-of-Distribution Detection and Transfer Learning](https://mlsanity.com/paper/2502.10691/) (2025)
- [Mysteries of the Deep: Role of Intermediate Representations in Out of Distribution Detection](https://mlsanity.com/paper/2510.05782/) (2025)
- [Multi-layer Aggregation as a key to feature-based OOD detection](https://mlsanity.com/paper/2307.15647/) (2023)
- [Leveraging Intermediate Representations for Better Out-of-Distribution Detection](https://mlsanity.com/paper/2502.12849/) (2025)
- [EOOD: Entropy-based Out-of-distribution Detection](https://mlsanity.com/paper/2504.03342/) (2025)
- [FisherMask: Enhancing Neural Network Labeling Efficiency in Image Classification Using Fisher Information](https://mlsanity.com/paper/2411.05752/) (2024)
- [LPLgrad: Optimizing Active Learning Through Gradient Norm Sample Selection and Auxiliary Model Training](https://mlsanity.com/paper/2411.15217/) (2024)

## Abstract (arXiv)

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.