Blind Spot Navigation in Large Language Model Reasoning with Thought Space Explorer
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914119890763776 |
|---|---|
| author | Zhang, Jinghan Mo, Fengran Weerasooriya, Tharindu Cyril Ye, Xinyue Wang, Dongjie Fu, Yanjie Liu, Kunpeng |
| author_facet | Zhang, Jinghan Mo, Fengran Weerasooriya, Tharindu Cyril Ye, Xinyue Wang, Dongjie Fu, Yanjie Liu, Kunpeng |
| contents | Large language models have shown strong reasoning capabilities through chain-structured methods such as Chain-of-Thought. Recent studies optimize thought structures by generating parallel or tree-like structures, switching between long and short reasoning modes, or aligning reasoning steps with task performance. However, these approaches mainly rely on previously generated logical directions of the chains, which ignore the unexplored regions of the solution space. Such a phenomenon is defined as blind spots, which limit the diversity and effectiveness of the reasoning process. To this end, we propose the ``Thought Space Explorer'' (TSE), a framework for navigating and expanding thought structures to overcome blind spots in LLM reasoning. Our TSE first identifies key nodes with high impact, then generates new nodes by integrating information from multiple chains. Finally, it extends new branches through connection strategies. We conduct a series of experiments on math and QA benchmarks. Compared with existing baseline methods, TSE improves the accuracy of both the final answer and intermediate reasoning steps, while maintaining a better effectiveness-efficiency trade-off for practical deployment. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2410_24155 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Blind Spot Navigation in Large Language Model Reasoning with Thought Space Explorer Zhang, Jinghan Mo, Fengran Weerasooriya, Tharindu Cyril Ye, Xinyue Wang, Dongjie Fu, Yanjie Liu, Kunpeng Computation and Language Large language models have shown strong reasoning capabilities through chain-structured methods such as Chain-of-Thought. Recent studies optimize thought structures by generating parallel or tree-like structures, switching between long and short reasoning modes, or aligning reasoning steps with task performance. However, these approaches mainly rely on previously generated logical directions of the chains, which ignore the unexplored regions of the solution space. Such a phenomenon is defined as blind spots, which limit the diversity and effectiveness of the reasoning process. To this end, we propose the ``Thought Space Explorer'' (TSE), a framework for navigating and expanding thought structures to overcome blind spots in LLM reasoning. Our TSE first identifies key nodes with high impact, then generates new nodes by integrating information from multiple chains. Finally, it extends new branches through connection strategies. We conduct a series of experiments on math and QA benchmarks. Compared with existing baseline methods, TSE improves the accuracy of both the final answer and intermediate reasoning steps, while maintaining a better effectiveness-efficiency trade-off for practical deployment. |
| title | Blind Spot Navigation in Large Language Model Reasoning with Thought Space Explorer |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2410.24155 |