ALLURE: Auditing and Improving LLM-based Evaluation of Text using Iterative In-Context-Learning
This addresses the need for more reliable LLM evaluators in domains like medical summarization and education, though it appears incremental as an enhancement to existing in-context learning approaches.
The paper tackles the problem of failure modes in LLM-based text evaluation by introducing ALLURE, a method that audits and improves LLM evaluators through iterative in-context learning, ultimately reducing reliance on human annotators.
From grading papers to summarizing medical documents, large language models (LLMs) are evermore used for evaluation of text generated by humans and AI alike. However, despite their extensive utility, LLMs exhibit distinct failure modes, necessitating a thorough audit and improvement of their text evaluation capabilities. Here we introduce ALLURE, a systematic approach to Auditing Large Language Models Understanding and Reasoning Errors. ALLURE involves comparing LLM-generated evaluations with annotated data, and iteratively incorporating instances of significant deviation into the evaluator, which leverages in-context learning (ICL) to enhance and improve robust evaluation of text by LLMs. Through this iterative process, we refine the performance of the evaluator LLM, ultimately reducing reliance on human annotators in the evaluation process. We anticipate ALLURE to serve diverse applications of LLMs in various domains related to evaluation of textual data, such as medical summarization, education, and and productivity.