
Teaching large language models to use criteria for feedback generation is essential for improving task performance and aligning with human values.


coremsg

Teaching-Large-Language-Models-to-Use-Criteria-for-Feedback-Generation


Teaching Large Language Models to Use Criteria for Feedback Generation



大規模言語モデル（LLMs）に包括的な基準を使用させて、自然言語フィードバックを提供するためのフレームワークを提案します。



LLMCRIT: Enabling Large Language Models to Use Comprehensive Criteria for Feedback Generation
