HypoEval: Hypothesis-Guided Evaluation for Natural Language Generation
This provides a more reliable and interpretable automated evaluation framework for natural language generation, addressing a bottleneck in LLM-as-a-judge methods, though it is incremental in improving existing approaches.
The paper tackles the problem of low alignment and lack of interpretability in LLM-based evaluation of natural language generation by proposing HypoEval, a hypothesis-guided framework that uses a small set of human evaluations to generate detailed rubrics and combines LLM scores across dimensions. With only 30 human evaluations, it achieves state-of-the-art performance, outperforming G-Eval by 11.86% and fine-tuned Llama-3.1-8B-Instruct by 11.95% in correlation with human judgments.
Large language models (LLMs) have demonstrated great potential for automating the evaluation of natural language generation. Previous frameworks of LLM-as-a-judge fall short in two ways: they either use zero-shot setting without consulting any human input, which leads to low alignment, or fine-tune LLMs on labeled data, which requires a non-trivial number of samples. Moreover, previous methods often provide little reasoning behind automated evaluations. In this paper, we propose HypoEval, Hypothesis-guided Evaluation framework, which first uses a small corpus of human evaluations to generate more detailed rubrics for human judgments and then incorporates a checklist-like approach to combine LLM's assigned scores on each decomposed dimension to acquire overall scores. With only 30 human evaluations, HypoEval achieves state-of-the-art performance in alignment with both human rankings (Spearman correlation) and human scores (Pearson correlation), on average outperforming G-Eval by 11.86% and fine-tuned Llama-3.1-8B-Instruct with at least 3 times more human evaluations by 11.95%. Furthermore, we conduct systematic studies to assess the robustness of HypoEval, highlighting its effectiveness as a reliable and interpretable automated evaluation framework.