EvoWiki: Evaluating LLMs on Evolving Knowledge

TL;DR

EvoWiki is a dynamic benchmark dataset designed to evaluate how well large language models adapt to changing knowledge over time, addressing limitations of static benchmarks and highlighting the challenges and potential solutions.

Contribution

We introduce EvoWiki, an auto-updatable dataset that captures knowledge evolution, enabling precise evaluation of LLMs' adaptation to changing information.

Findings

Current LLMs often provide outdated or incorrect responses to evolved knowledge.

RAG and CL methods show a synergistic effect in improving adaptation.

EvoWiki offers a robust benchmark for future research on knowledge evolution in LLMs.

Abstract

Knowledge utilization is a critical aspect of LLMs, and understanding how they adapt to evolving knowledge is essential for their effective deployment. However, existing benchmarks are predominantly static, failing to capture the evolving nature of LLMs and knowledge, leading to inaccuracies and vulnerabilities such as contamination. In this paper, we introduce EvoWiki, an evolving dataset designed to reflect knowledge evolution by categorizing information into stable, evolved, and uncharted states. EvoWiki is fully auto-updatable, enabling precise evaluation of continuously changing knowledge and newly released LLMs. Through experiments with Retrieval-Augmented Generation (RAG) and Contunual Learning (CL), we evaluate how effectively LLMs adapt to evolving knowledge. Our results indicate that current models often struggle with evolved knowledge, frequently providing outdated or incorrect responses. Moreover, the dataset highlights a synergistic effect between RAG and CL, demonstrating their potential to better adapt to evolving knowledge. EvoWiki provides a robust benchmark for advancing future research on the knowledge evolution capabilities of large language models.