ReAgent: Reversible Multi-Agent Reasoning for Knowledge-Enhanced Multi-Hop QA
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866913865696018432 |
|---|---|
| author | Zhao, Xinjie Gao, Fan Song, Xingyu Chen, Yingjian Yang, Rui Fu, Yanran Wang, Yuyang Iwasawa, Yusuke Matsuo, Yutaka Li, Irene |
| author_facet | Zhao, Xinjie Gao, Fan Song, Xingyu Chen, Yingjian Yang, Rui Fu, Yanran Wang, Yuyang Iwasawa, Yusuke Matsuo, Yutaka Li, Irene |
| contents | Recent advances in large language models (LLMs) have significantly improved multi-hop question answering (QA) through direct Chain-of-Thought (CoT) reasoning. However, the irreversible nature of CoT leads to error accumulation, making it challenging to correct mistakes in multi-hop reasoning. This paper introduces ReAgent: a Reversible multi-Agent collaborative framework augmented with explicit backtracking mechanisms, enabling reversible multi-hop reasoning. By incorporating text-based retrieval, information aggregation and validation, our system can detect and correct errors mid-reasoning, leading to more robust and interpretable QA outcomes. The framework and experiments serve as a foundation for future work on error-tolerant QA systems. Empirical evaluations across three benchmarks indicate ReAgent's efficacy, yielding average about 6\% improvements against baseline models. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_06951 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | ReAgent: Reversible Multi-Agent Reasoning for Knowledge-Enhanced Multi-Hop QA Zhao, Xinjie Gao, Fan Song, Xingyu Chen, Yingjian Yang, Rui Fu, Yanran Wang, Yuyang Iwasawa, Yusuke Matsuo, Yutaka Li, Irene Artificial Intelligence Recent advances in large language models (LLMs) have significantly improved multi-hop question answering (QA) through direct Chain-of-Thought (CoT) reasoning. However, the irreversible nature of CoT leads to error accumulation, making it challenging to correct mistakes in multi-hop reasoning. This paper introduces ReAgent: a Reversible multi-Agent collaborative framework augmented with explicit backtracking mechanisms, enabling reversible multi-hop reasoning. By incorporating text-based retrieval, information aggregation and validation, our system can detect and correct errors mid-reasoning, leading to more robust and interpretable QA outcomes. The framework and experiments serve as a foundation for future work on error-tolerant QA systems. Empirical evaluations across three benchmarks indicate ReAgent's efficacy, yielding average about 6\% improvements against baseline models. |
| title | ReAgent: Reversible Multi-Agent Reasoning for Knowledge-Enhanced Multi-Hop QA |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2503.06951 |