作者位次 #4/9ACL 2026Findings· Findings of ACL
Exploration-Exploitation Reshaping towards Efficient Reasoning for Large Language Models
Yufeng Shi · Weilin Luo · Yuxiang Zhang · Zongmeng Zhang · Haoyang Liu · Yubing Wang · Bin Wang · Wengang Zhou · Houqiang Li
摘要
While excelling at solving complex problems, Large Reasoning Models (LRMs) are still constrained by the overthinking issue. Most current studies rely on reward shaping in Reinforcement Learning (RL) to shorten the Chain-of-Thought (CoT) of LRMs, remaining sample-inefficient and non-robust due to...