AdaMem: Adaptive User-Centric Memory for Long-Horizon Dialogue Agents
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_ | 1866910177101348864 |
|---|---|
| author | Yan, Shannan Ni, Jingchen Zheng, Leqi Zhang, Jiajun Wu, Peixi Yin, Dacheng Lyu, Jing Yuan, Chun Rao, Fengyun |
| author_facet | Yan, Shannan Ni, Jingchen Zheng, Leqi Zhang, Jiajun Wu, Peixi Yin, Dacheng Lyu, Jing Yuan, Chun Rao, Fengyun |
| contents | Large language model (LLM) agents increasingly rely on external memory to support long-horizon interaction, personalized assistance, and multi-step reasoning. However, existing memory systems still face three core challenges: they often rely too heavily on semantic similarity, which can miss evidence crucial for user-centric understanding; they frequently store related experiences as isolated fragments, weakening temporal and causal coherence; and they typically use static memory granularities that do not adapt well to the requirements of different questions. We propose AdaMem, an adaptive user-centric memory framework for long-horizon dialogue agents. AdaMem organizes dialogue history into working, episodic, persona, and graph memories, enabling the system to preserve recent context, structured long-term experiences, stable user traits, and relation-aware connections within a unified framework. At inference time, AdaMem first resolves the target participant, then builds a question-conditioned retrieval route that combines semantic retrieval with relation-aware graph expansion only when needed, and finally produces the answer through a role-specialized pipeline for evidence synthesis and response generation. We evaluate AdaMem on the LoCoMo and PERSONAMEM benchmarks for long-horizon reasoning and user modeling. Experimental results show that AdaMem achieves state-of-the-art performance on both benchmarks. The code will be released upon acceptance. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_16496 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | AdaMem: Adaptive User-Centric Memory for Long-Horizon Dialogue Agents Yan, Shannan Ni, Jingchen Zheng, Leqi Zhang, Jiajun Wu, Peixi Yin, Dacheng Lyu, Jing Yuan, Chun Rao, Fengyun Computation and Language Large language model (LLM) agents increasingly rely on external memory to support long-horizon interaction, personalized assistance, and multi-step reasoning. However, existing memory systems still face three core challenges: they often rely too heavily on semantic similarity, which can miss evidence crucial for user-centric understanding; they frequently store related experiences as isolated fragments, weakening temporal and causal coherence; and they typically use static memory granularities that do not adapt well to the requirements of different questions. We propose AdaMem, an adaptive user-centric memory framework for long-horizon dialogue agents. AdaMem organizes dialogue history into working, episodic, persona, and graph memories, enabling the system to preserve recent context, structured long-term experiences, stable user traits, and relation-aware connections within a unified framework. At inference time, AdaMem first resolves the target participant, then builds a question-conditioned retrieval route that combines semantic retrieval with relation-aware graph expansion only when needed, and finally produces the answer through a role-specialized pipeline for evidence synthesis and response generation. We evaluate AdaMem on the LoCoMo and PERSONAMEM benchmarks for long-horizon reasoning and user modeling. Experimental results show that AdaMem achieves state-of-the-art performance on both benchmarks. The code will be released upon acceptance. |
| title | AdaMem: Adaptive User-Centric Memory for Long-Horizon Dialogue Agents |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2603.16496 |