A Survey on Out-of-Distribution Detection in NLP

Hao Lang, Yinhe Zheng, Yixuan Li, Jian Sun, Fei Huang +1 more
5/5/2023
cs.CLcs.AIcs.LG

No official code found. Best implementation: ca1man-2022/Awesome-GOOD-Detection (13 stars, last commit 2026-02-11).

Abstract

Out-of-distribution (OOD) detection is essential for the reliable and safe deployment of machine learning systems in the real world. Great progress has been made over the past years. This paper presents the first review of recent advances in OOD detection with a particular focus on natural language processing approaches. First, we provide a formal definition of OOD detection and discuss several related fields. We then categorize recent algorithms into three classes according to the data they used: (1) OOD data available, (2) OOD data unavailable + in-distribution (ID) label available, and (3) OOD data unavailable + ID label unavailable. Third, we introduce datasets, applications, and metrics. Finally, we summarize existing work and present potential future research topics.

View on arXivView PDF

Code Implementations(1)

ca1man-2022/Awesome-GOOD-DetectionCommunity71%

A survey of graph out-of-distribution detection

131Jun 20, 20248 months ago

Similar Papers

Out-of-Distribution Detection: A Task-Oriented Survey of Recent Advances

88% match1 repo1732024

FLatS: Principled Out-of-Distribution Detection with Feature-Based Likelihood Ratio Score

86% match1 repo42023

When and How Does In-Distribution Label Help Out-of-Distribution Detection?

85% match1 repo42024

Unified Out-Of-Distribution Detection: A Model-Specific Perspective

85% match2023

Is Fine-tuning Needed? Pre-trained Language Models Are Near Perfect for Out-of-Domain Detection

85% match2023

A Closer Look at the Learnability of Out-of-Distribution (OOD) Detection

85% match2025

Cite this paper

@article{lang2023survey,
  title  = {A Survey on Out-of-Distribution Detection in NLP},
  author = {Hao Lang and Yinhe Zheng and Yixuan Li and Jian Sun and Fei Huang and Yongbin Li},
  year   = {2023},
  eprint = {2305.03236},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2305.03236v2}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2305.03236.svg)](https://codeofpaper.com/paper/2305.03236)

Discussion