O-Mem: Omni Memory System for Personalized, Long Horizon, Self-Evolving Agents
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908704176078848 |
|---|---|
| author | Wang, Piaohong Tian, Motong Li, Jiaxian Liang, Yuan Wang, Yuqing Chen, Qianben Wang, Tiannan Lu, Zhicong Ma, Jiawei Jiang, Yuchen Eleanor Zhou, Wangchunshu |
| author_facet | Wang, Piaohong Tian, Motong Li, Jiaxian Liang, Yuan Wang, Yuqing Chen, Qianben Wang, Tiannan Lu, Zhicong Ma, Jiawei Jiang, Yuchen Eleanor Zhou, Wangchunshu |
| contents | Recent advancements in LLM-powered agents have demonstrated significant potential in generating human-like responses; however, they continue to face challenges in maintaining long-term interactions within complex environments, primarily due to limitations in contextual consistency and dynamic personalization. Existing memory systems often depend on semantic grouping prior to retrieval, which can overlook semantically irrelevant yet critical user information and introduce retrieval noise. In this report, we propose the initial design of O-Mem, a novel memory framework based on active user profiling that dynamically extracts and updates user characteristics and event records from their proactive interactions with agents. O-Mem supports hierarchical retrieval of persona attributes and topic-related context, enabling more adaptive and coherent personalized responses. O-Mem achieves 51.67% on the public LoCoMo benchmark, a nearly 3% improvement upon LangMem,the previous state-of-the-art, and it achieves 62.99% on PERSONAMEM, a 3.5% improvement upon A-Mem,the previous state-of-the-art. O-Mem also boosts token and interaction response time efficiency compared to previous memory frameworks. Our work opens up promising directions for developing efficient and human-like personalized AI assistants in the future. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_13593 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | O-Mem: Omni Memory System for Personalized, Long Horizon, Self-Evolving Agents Wang, Piaohong Tian, Motong Li, Jiaxian Liang, Yuan Wang, Yuqing Chen, Qianben Wang, Tiannan Lu, Zhicong Ma, Jiawei Jiang, Yuchen Eleanor Zhou, Wangchunshu Computation and Language Recent advancements in LLM-powered agents have demonstrated significant potential in generating human-like responses; however, they continue to face challenges in maintaining long-term interactions within complex environments, primarily due to limitations in contextual consistency and dynamic personalization. Existing memory systems often depend on semantic grouping prior to retrieval, which can overlook semantically irrelevant yet critical user information and introduce retrieval noise. In this report, we propose the initial design of O-Mem, a novel memory framework based on active user profiling that dynamically extracts and updates user characteristics and event records from their proactive interactions with agents. O-Mem supports hierarchical retrieval of persona attributes and topic-related context, enabling more adaptive and coherent personalized responses. O-Mem achieves 51.67% on the public LoCoMo benchmark, a nearly 3% improvement upon LangMem,the previous state-of-the-art, and it achieves 62.99% on PERSONAMEM, a 3.5% improvement upon A-Mem,the previous state-of-the-art. O-Mem also boosts token and interaction response time efficiency compared to previous memory frameworks. Our work opens up promising directions for developing efficient and human-like personalized AI assistants in the future. |
| title | O-Mem: Omni Memory System for Personalized, Long Horizon, Self-Evolving Agents |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2511.13593 |