paper-with-me

홈 › Papers

Mitigating Neural Network Overconfidence with Logit Normalization

2022-05-19 · Hongxin Wei, Renchunzi Xie, Hao Cheng, Lei Feng, Bo An, Yixuan Li

Detecting out-of-distribution inputs is critical for safe deployment of machine learning models in the real world. However, neural networks are known to suffer from the overconfidence issue, where they produce abnormally high confidence for both in- and out-of-distribution inputs. In this work, we show that this issue can be mitigated through Logit Normalization (LogitNorm) -- a simple fix to the cross-entropy loss -- by enforcing a constant vector norm on the logits in training. Our method is motivated by the analysis that the norm of the logit keeps increasing during training, leading to overconfident output. Our key idea behind LogitNorm is thus to decouple the influence of output's norm during network optimization. Trained with LogitNorm, neural networks produce highly distinguishable confidence scores between in- and out-of-distribution data. Extensive experiments demonstrate the superiority of LogitNorm, reducing the average FPR95 by up to 42.30% on common benchmarks.

📄 PDF Abstract BibTeX arXiv:2205.09310

Code (2)

hongxin001/logitnorm_ood 공식 구현 pytorch
hongxin001/open-sampling pytorch

Similar Papers 제목 키워드 기반

MANO: Exploiting Matrix Norm for Unsupervised Accuracy Estimation Under Distribution Shifts

2024-05-29 · Renchunzi Xie, Ambroise Odonnat, Vasilii Feofanov, Weijian Deng 외

Leveraging the models' outputs, specifically the logits, is a common approach to estimating the test accuracy of a pre-trained neural network on out-of-distribution (OOD) samples without requiring access to the correspon…

Computational Efficiency

Typicalness-Aware Learning for Failure Detection

2024-11-04 · Yijun Liu, Jiequan Cui, Zhuotao Tian, Senqiao Yang 외

Deep neural networks (DNNs) often suffer from the overconfidence issue, where incorrect predictions are made with high confidence scores, hindering the applications in critical systems. In this paper, we propose a novel …

Similarity-as-Evidence: Calibrating Overconfident VLMs for Interpretable and Label-Efficient Medical Active Learning

2026-02-21 · Zhuofan Xie, Zishan Lin, Jinliang Lin, Jie Qi 외 arxiv

Active Learning (AL) reduces annotation costs in medical imaging by selecting only the most informative samples for labeling, but suffers from cold-start when labeled data are scarce. Vision-Language Models (VLMs) addres…

Active Learning

Enhancing Out-of-Distribution Detection with Extended Logit Normalization

2025-04-15 · Yifan Ding, Xixi Liu, Jonas Unger, Gabriel Eilertsen

Out-of-distribution (OOD) detection is essential for the safe deployment of machine learning models. Recent advances have explored improved classification losses and representation learning strategies to enhance OOD dete…

Out-of-Distribution DetectionOut of Distribution (OOD) DetectionRepresentation Learning

Uncalibrated Reasoning: GRPO Induces Overconfidence for Stochastic Outcomes

2025-08-15 · Michael Bereket, Jure Leskovec arxiv

Reinforcement learning (RL) has proven remarkably effective at improving the accuracy of language models in verifiable and deterministic domains like mathematics. Here, we examine if current RL methods are also effective…

Reinforcement Learning