作者位次 #6/9ACL 2026Findings· Findings of ACL
MT - OSC : Path for LLM s that Get Lost in Multi-Turn Conversation
Jyotika Singh · Fang Tu · Miguel Ballesteros · Weiyi Sun · Sandip Ghoshal · Michelle Yuan · Yassine Benajiba · Sujith Ravi · Dan Roth
摘要
Large language models (LLMs) suffer significant performance degradation when user instructions and context are distributed over multiple conversational turns, yet multi-turn (MT) interactions dominate chat interfaces. The routine approach of appending full chat history to prompts rapidly exhausts...