Out-of-Distribution Detection using Synthetic Data Generation

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Abbas, Momin, Azmat, Muneeza, Horesh, Raya, Yurochkin, Mikhail
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866916983700717568
author Abbas, Momin
Azmat, Muneeza
Horesh, Raya
Yurochkin, Mikhail
author_facet Abbas, Momin
Azmat, Muneeza
Horesh, Raya
Yurochkin, Mikhail
contents Distinguishing in- and out-of-distribution (OOD) inputs is crucial for reliable deployment of classification systems. However, OOD data is typically unavailable or difficult to collect, posing a significant challenge for accurate OOD detection. In this work, we present a method that harnesses the generative capabilities of Large Language Models (LLMs) to create high-quality synthetic OOD proxies, eliminating the dependency on any external OOD data source. We study the efficacy of our method on classical text classification tasks such as toxicity detection and sentiment classification as well as classification tasks arising in LLM development and deployment, such as training a reward model for RLHF and detecting misaligned generations. Extensive experiments on nine InD-OOD dataset pairs and various model sizes show that our approach dramatically lowers false positive rates (achieving a perfect zero in some cases) while maintaining high accuracy on in-distribution tasks, outperforming baseline methods by a significant margin.
format Preprint
id arxiv_https___arxiv_org_abs_2502_03323
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Out-of-Distribution Detection using Synthetic Data Generation
Abbas, Momin
Azmat, Muneeza
Horesh, Raya
Yurochkin, Mikhail
Computation and Language
Artificial Intelligence
Machine Learning
Distinguishing in- and out-of-distribution (OOD) inputs is crucial for reliable deployment of classification systems. However, OOD data is typically unavailable or difficult to collect, posing a significant challenge for accurate OOD detection. In this work, we present a method that harnesses the generative capabilities of Large Language Models (LLMs) to create high-quality synthetic OOD proxies, eliminating the dependency on any external OOD data source. We study the efficacy of our method on classical text classification tasks such as toxicity detection and sentiment classification as well as classification tasks arising in LLM development and deployment, such as training a reward model for RLHF and detecting misaligned generations. Extensive experiments on nine InD-OOD dataset pairs and various model sizes show that our approach dramatically lowers false positive rates (achieving a perfect zero in some cases) while maintaining high accuracy on in-distribution tasks, outperforming baseline methods by a significant margin.
title Out-of-Distribution Detection using Synthetic Data Generation
topic Computation and Language
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2502.03323