DTS: Enhancing Large Reasoning Models via Decoding Tree Sketching
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866908810839326720 |
|---|---|
| author | Xu, Zicheng Lou, Xiuyi Wang, Guanchu Chuang, Yu-Neng Luo, Feng Zheng, Guangyao Szalay, Alexander S. Liu, Zirui Braverman, Vladimir |
| author_facet | Xu, Zicheng Lou, Xiuyi Wang, Guanchu Chuang, Yu-Neng Luo, Feng Zheng, Guangyao Szalay, Alexander S. Liu, Zirui Braverman, Vladimir |
| contents | Large Reasoning Models (LRMs) achieve remarkable inference-time improvements through parallel thinking. However, existing methods rely on redundant sampling of reasoning trajectories, failing to effectively explore the reasoning space to uncover high-quality solutions. To address these limitations, we propose Decoding Tree Sketching (DTS), a plug-and-play decoding framework for structural multi-trajectory exploration and reasoning selection. For reasoning exploration, DTS sketches a backbone tree of the reasoning space by selectively branching at decision tokens. For reasoning selection, guided by length-accuracy anti-correlation, DTS designs an early termination to prioritize short and reliable trajectories during decoding. Experimental results across four LRMs and datasets demonstrate that DTS significantly enhances accuracy by 14% and reduces repetitive generation by 8% on average. Notably, DTS enables smaller models to outperform larger models with 10$\times$ the size, highlighting its potential to strengthen reasoning capabilities. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_00640 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | DTS: Enhancing Large Reasoning Models via Decoding Tree Sketching Xu, Zicheng Lou, Xiuyi Wang, Guanchu Chuang, Yu-Neng Luo, Feng Zheng, Guangyao Szalay, Alexander S. Liu, Zirui Braverman, Vladimir Artificial Intelligence Computation and Language Machine Learning Large Reasoning Models (LRMs) achieve remarkable inference-time improvements through parallel thinking. However, existing methods rely on redundant sampling of reasoning trajectories, failing to effectively explore the reasoning space to uncover high-quality solutions. To address these limitations, we propose Decoding Tree Sketching (DTS), a plug-and-play decoding framework for structural multi-trajectory exploration and reasoning selection. For reasoning exploration, DTS sketches a backbone tree of the reasoning space by selectively branching at decision tokens. For reasoning selection, guided by length-accuracy anti-correlation, DTS designs an early termination to prioritize short and reliable trajectories during decoding. Experimental results across four LRMs and datasets demonstrate that DTS significantly enhances accuracy by 14% and reduces repetitive generation by 8% on average. Notably, DTS enables smaller models to outperform larger models with 10$\times$ the size, highlighting its potential to strengthen reasoning capabilities. |
| title | DTS: Enhancing Large Reasoning Models via Decoding Tree Sketching |
| topic | Artificial Intelligence Computation and Language Machine Learning |
| url | https://arxiv.org/abs/2511.00640 |