1 paper
Wenchang Lei, Ping Zou, Yue Wang +2
Large language models (LLMs) exhibit strong semantic understanding, yet struggle when user instructions involve ambiguous or conceptually misaligned terms. We propose the Language…