一作ACL 2026Findings· Findings of ACL
MAPLE : Multi-Aspect Panels of LLM Evaluators for Open-Ended Questions
Michinori Jinji · Kyohei Atarashi · Koh Takeuchi · Hisashi Kashima
摘要
LLM-as-a-Judge, which uses LLMs to evaluate responses to open-ended questions, has seen significant growth in recent years. It has been adopted as a scalable alternative to manual human evaluation, such as crowdsourcing, which is often time-consuming and costly. However, the discrepancy between L...