一作ACL 2026Findings· Findings of ACL
Efficient Reasoning for LLM s through Speculative Chain-of-Thought
Jikai Wang · Juntao Li · Jianye Hou · Yan Bowen · Lijun Wu · Min Zhang
摘要
Large reasoning language models such as OpenAI-o1 and Deepseek-R1 have recently attracted widespread attention due to their impressive task-solving abilities. However, the enormous model size and the generation of lengthy thought chains introduce significant reasoning costs and response latency....