Thought Manipulation: External Thought Can Be Efficient for Large Reasoning Models
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866912517926682624 |
|---|---|
| author | Liu, Yule Zheng, Jingyi Sun, Zhen Peng, Zifan Dong, Wenhan Sha, Zeyang Cui, Shiwen Wang, Weiqiang He, Xinlei |
| author_facet | Liu, Yule Zheng, Jingyi Sun, Zhen Peng, Zifan Dong, Wenhan Sha, Zeyang Cui, Shiwen Wang, Weiqiang He, Xinlei |
| contents | Recent advancements in large reasoning models (LRMs) have demonstrated the effectiveness of scaling test-time computation to enhance reasoning capabilities on various tasks. However, LRMs often suffer from an ``overthinking'' problem, where the model generates excessively redundant reasoning steps with limited performance gains. In this work, we empirically reveal an important characteristic of LRM behaviors that placing external CoTs generated by smaller models between the thinking token (\texttt{<think>} and \texttt{</think>}) can effectively manipulate the model to generate fewer thoughts. Building on this finding, we propose a simple yet efficient pipeline, \Method, to enable LRMs to bypass unnecessary intermediate steps, thereby significantly reducing computational costs. We conduct extensive experiments to evaluate the utility and efficiency of \Method. For instance, when applied to QwQ-32B on the LiveBench/Code dataset, \Method keeps the original performance while reducing output token counts by approximately 30\%, with minimal overhead introduced by the CoT generator. Furthermore, we identify two suboptimal modes, blindly following flawed external thoughts and unnecessary rethinking, and show that simple mitigations, such as difficulty-aware fallbacks, can further improve performance. Overall, \Method offers a practical, general, and efficient way to optimize LRM inference, making powerful reasoning models more accessible and scalable for real-world applications. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_13626 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Thought Manipulation: External Thought Can Be Efficient for Large Reasoning Models Liu, Yule Zheng, Jingyi Sun, Zhen Peng, Zifan Dong, Wenhan Sha, Zeyang Cui, Shiwen Wang, Weiqiang He, Xinlei Computation and Language Artificial Intelligence Recent advancements in large reasoning models (LRMs) have demonstrated the effectiveness of scaling test-time computation to enhance reasoning capabilities on various tasks. However, LRMs often suffer from an ``overthinking'' problem, where the model generates excessively redundant reasoning steps with limited performance gains. In this work, we empirically reveal an important characteristic of LRM behaviors that placing external CoTs generated by smaller models between the thinking token (\texttt{<think>} and \texttt{</think>}) can effectively manipulate the model to generate fewer thoughts. Building on this finding, we propose a simple yet efficient pipeline, \Method, to enable LRMs to bypass unnecessary intermediate steps, thereby significantly reducing computational costs. We conduct extensive experiments to evaluate the utility and efficiency of \Method. For instance, when applied to QwQ-32B on the LiveBench/Code dataset, \Method keeps the original performance while reducing output token counts by approximately 30\%, with minimal overhead introduced by the CoT generator. Furthermore, we identify two suboptimal modes, blindly following flawed external thoughts and unnecessary rethinking, and show that simple mitigations, such as difficulty-aware fallbacks, can further improve performance. Overall, \Method offers a practical, general, and efficient way to optimize LRM inference, making powerful reasoning models more accessible and scalable for real-world applications. |
| title | Thought Manipulation: External Thought Can Be Efficient for Large Reasoning Models |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2504.13626 |