MemSearch-o1: Empowering Large Language Models with Reasoning-Aligned Memory Growth in Agentic Search
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913115169357824 |
|---|---|
| author | Zhang, Sheng Li, Junyi Zhang, Yingyi Jia, Pengyue Wang, Yichao Qian, Xiaowei Zhang, Wenlin Wang, Maolin Liu, Yong Zhao, Xiangyu |
| author_facet | Zhang, Sheng Li, Junyi Zhang, Yingyi Jia, Pengyue Wang, Yichao Qian, Xiaowei Zhang, Wenlin Wang, Maolin Liu, Yong Zhao, Xiangyu |
| contents | Recent advances in large language models (LLMs) have scaled the potential for reasoning and agentic search, wherein models autonomously plan, retrieve, and reason over external knowledge to answer complex queries. However, the iterative think-search loop accumulates long system memories, leading to memory dilution problem. In addition, existing memory management methods struggle to capture fine-grained semantic relations between queries and documents and often lose substantial information. Therefore, we propose MemSearch-o1, an agentic search framework built on reasoning-aligned memory growth and retracing. MemSearch-o1 dynamically grows fine-grained memory fragments from memory seed tokens from the queries, then retraces and deeply refines the memory via a contribution function, and finally reorganizes a globally connected memory path. This shifts memory management from stream-like concatenation to structured, token-level growth with path-based reasoning. Experiments on eight benchmark datasets show that MemSearch-o1 substantially mitigates memory dilution, and more effectively activates the reasoning potential of diverse LLMs, establishing a solid foundation for memory-aware agentic intelligence. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_17265 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | MemSearch-o1: Empowering Large Language Models with Reasoning-Aligned Memory Growth in Agentic Search Zhang, Sheng Li, Junyi Zhang, Yingyi Jia, Pengyue Wang, Yichao Qian, Xiaowei Zhang, Wenlin Wang, Maolin Liu, Yong Zhao, Xiangyu Information Retrieval Recent advances in large language models (LLMs) have scaled the potential for reasoning and agentic search, wherein models autonomously plan, retrieve, and reason over external knowledge to answer complex queries. However, the iterative think-search loop accumulates long system memories, leading to memory dilution problem. In addition, existing memory management methods struggle to capture fine-grained semantic relations between queries and documents and often lose substantial information. Therefore, we propose MemSearch-o1, an agentic search framework built on reasoning-aligned memory growth and retracing. MemSearch-o1 dynamically grows fine-grained memory fragments from memory seed tokens from the queries, then retraces and deeply refines the memory via a contribution function, and finally reorganizes a globally connected memory path. This shifts memory management from stream-like concatenation to structured, token-level growth with path-based reasoning. Experiments on eight benchmark datasets show that MemSearch-o1 substantially mitigates memory dilution, and more effectively activates the reasoning potential of diverse LLMs, establishing a solid foundation for memory-aware agentic intelligence. |
| title | MemSearch-o1: Empowering Large Language Models with Reasoning-Aligned Memory Growth in Agentic Search |
| topic | Information Retrieval |
| url | https://arxiv.org/abs/2604.17265 |