Know Your Limits: A Survey of Abstention in Large Language Models

TL;DR

This survey reviews the current state of abstention in large language models, highlighting methods, benchmarks, and evaluation metrics to improve safety and reliability, and discusses future research directions.

Contribution

It introduces a comprehensive framework for analyzing abstention in LLMs and organizes existing literature within this structure, identifying gaps and future opportunities.

Findings

Abstention can reduce hallucinations and improve safety in LLMs.

Current methods vary in effectiveness and scope.

Future research should focus on meta-capabilities and context-specific abstention.

Abstract

Abstention, the refusal of large language models (LLMs) to provide an answer, is increasingly recognized for its potential to mitigate hallucinations and enhance safety in LLM systems. In this survey, we introduce a framework to examine abstention from three perspectives: the query, the model, and human values. We organize the literature on abstention methods, benchmarks, and evaluation metrics using this framework, and discuss merits and limitations of prior work. We further identify and motivate areas for future research, such as whether abstention can be achieved as a meta-capability that transcends specific tasks or domains, and opportunities to optimize abstention abilities in specific contexts. In doing so, we aim to broaden the scope and impact of abstention methodologies in AI systems.