Oddballness: universal anomaly detection with language models

Filip Graliński, Ryszard Staruch, Krzysztof Jurkiewicz
9/4/2024
cs.CL

No confident code implementation found for this paper.

Abstract

We present a new method to detect anomalies in texts (in general: in sequences of any data), using language models, in a totally unsupervised manner. The method considers probabilities (likelihoods) generated by a language model, but instead of focusing on low-likelihood tokens, it considers a new metric introduced in this paper: oddballness. Oddballness measures how ``strange'' a given token is according to the language model. We demonstrate in grammatical error detection tasks (a specific case of text anomaly detection) that oddballness is better than just considering low-likelihood events, if a totally unsupervised setup is assumed.

View on arXivView PDF

Code Implementations

No confident code match yet

We couldn't find an author-owned or strongly-evidenced community implementation for this paper. Any repos shown below are weak matches — verify before relying on them.

No code implementations found yet.

Know of an implementation? Let us know in the comments below!

Similar Papers

KKA: Improving Vision Anomaly Detection through Anomaly-related Knowledge from Large Language Models

84% match1 repo12025

Anomaly Detection in Human Language via Meta-Learning: A Few-Shot Approach

83% match2025

Mechanistic Anomaly Detection for "Quirky" Language Models

83% match2025

Large Language Models for Anomaly and Out-of-Distribution Detection: A Survey

83% match1 repo1242024

Random Word Data Augmentation with CLIP for Zero-Shot Anomaly Detection

83% match2023

Beyond a Single Perspective: Text Anomaly Detection with Multi-View Language Representations

83% match1 repo72026

Cite this paper

@article{graliski2024oddballness,
  title  = {Oddballness: universal anomaly detection with language models},
  author = {Filip Graliński and Ryszard Staruch and Krzysztof Jurkiewicz},
  year   = {2024},
  eprint = {2409.03046},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2409.03046v1}
}

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/2409.03046.svg)](https://codeofpaper.com/paper/2409.03046)

Discussion