TDA-RC: Task-Driven Alignment for Knowledge-Based Reasoning Chains in Large Language Models
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866913009235918848 |
|---|---|
| author | Zhang, Jiaquan Sun, Qigan Zhang, Chaoning Wang, Xudong Huang, Zhenzhen Zhou, Yitian Zheng, Pengcheng Tai, Chi-lok Andy Bae, Sung-Ho Ma, Zeyu Qin, Caiyan Guo, Jinyu Yang, Yang Shen, Hengtao |
| author_facet | Zhang, Jiaquan Sun, Qigan Zhang, Chaoning Wang, Xudong Huang, Zhenzhen Zhou, Yitian Zheng, Pengcheng Tai, Chi-lok Andy Bae, Sung-Ho Ma, Zeyu Qin, Caiyan Guo, Jinyu Yang, Yang Shen, Hengtao |
| contents | Enhancing the reasoning capability of large language models (LLMs) remains a core challenge in natural language processing. The Chain-of-Thought (CoT) paradigm dominates practical applications for its single-round efficiency, yet its reasoning chains often exhibit logical gaps. While multi-round paradigms like Graph-of-Thoughts (GoT), Tree-of-Thoughts (ToT), and Atom of Thought (AoT) achieve strong performance and reveal effective reasoning structures, their high cost limits practical use. To address this problem, this paper proposes a topology-based method for optimizing reasoning chains. The framework embeds essential topological patterns of effective reasoning into the lightweight CoT paradigm. Using persistent homology, we map CoT, ToT, and GoT into a unified topological space to quantify their structural features. On this basis, we design a unified optimization system: a Topological Optimization Agent diagnoses deviations in CoT chains from desirable topological characteristics and simultaneously generates targeted strategies to repair these structural deficiencies. Compared with multi-round reasoning methods like ToT and GoT, experiments on multiple datasets show that our approach offers a superior balance between reasoning accuracy and efficiency, showcasing a practical solution to ``single-round generation with multi-round intelligence''. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_04942 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | TDA-RC: Task-Driven Alignment for Knowledge-Based Reasoning Chains in Large Language Models Zhang, Jiaquan Sun, Qigan Zhang, Chaoning Wang, Xudong Huang, Zhenzhen Zhou, Yitian Zheng, Pengcheng Tai, Chi-lok Andy Bae, Sung-Ho Ma, Zeyu Qin, Caiyan Guo, Jinyu Yang, Yang Shen, Hengtao Computation and Language Artificial Intelligence Enhancing the reasoning capability of large language models (LLMs) remains a core challenge in natural language processing. The Chain-of-Thought (CoT) paradigm dominates practical applications for its single-round efficiency, yet its reasoning chains often exhibit logical gaps. While multi-round paradigms like Graph-of-Thoughts (GoT), Tree-of-Thoughts (ToT), and Atom of Thought (AoT) achieve strong performance and reveal effective reasoning structures, their high cost limits practical use. To address this problem, this paper proposes a topology-based method for optimizing reasoning chains. The framework embeds essential topological patterns of effective reasoning into the lightweight CoT paradigm. Using persistent homology, we map CoT, ToT, and GoT into a unified topological space to quantify their structural features. On this basis, we design a unified optimization system: a Topological Optimization Agent diagnoses deviations in CoT chains from desirable topological characteristics and simultaneously generates targeted strategies to repair these structural deficiencies. Compared with multi-round reasoning methods like ToT and GoT, experiments on multiple datasets show that our approach offers a superior balance between reasoning accuracy and efficiency, showcasing a practical solution to ``single-round generation with multi-round intelligence''. |
| title | TDA-RC: Task-Driven Alignment for Knowledge-Based Reasoning Chains in Large Language Models |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2604.04942 |