作者位次 #2/6ACL 2026Findings· Findings of ACL
How Do Answer Tokens Read Reasoning Traces? Self-Reading Patterns in Thinking LLM s for Quantitative Reasoning
Haoyang Chen · Yi Liu · Jianzhi Shao · Tao Zhang · Chengfu Huo · Wei Hu
摘要
Thinking LLMs produce reasoning traces before answering. Prior activation steering work mainly targets on shaping these traces. It remains less understood how answer tokens actually read and integrate the reasoning to produce reliable outcomes. Focusing on quantitative reasoning, we analyze the a...