Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2512.08870 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918280726315008 |
|---|---|
| author | Chen, Xiang Shi, Yuling Lan, Qizhen Qiu, Yuchao Wang, Min Gu, Xiaodong Yan, Yanfu |
| author_facet | Chen, Xiang Shi, Yuling Lan, Qizhen Qiu, Yuchao Wang, Min Gu, Xiaodong Yan, Yanfu |
| contents | LLM agents are widely deployed in complex interactive tasks, yet privacy constraints often preclude centralized optimization and co-evolution across dynamic environments. Despite the demonstrated success of Federated Learning (FL) on static datasets, its effectiveness in open-ended, self-evolving agent systems remains largely unexplored. In such settings, the direct application of standard FL is particularly challenging, as heterogeneous tasks and sparse, trajectory-level reward signals give rise to severe gradient instability, which undermines the global optimization process. To bridge this gap, we propose Fed-SE, a Federated Self-Evolution framework for LLM agents that establishes a local evolution-global aggregation paradigm. Locally, agents employ parameter-efficient fine-tuning on filtered, high-return trajectories to achieve stable gradient updates. Globally, Fed-SE aggregates updates within a low-rank subspace, reducing communication cost across clients. Experiments across five heterogeneous environments demonstrate that Fed-SE improves average task success rates by 10\% over the state-of-the-art FedIT, validating its effectiveness in cross-environment knowledge transfer under privacy constraints. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_08870 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Fed-SE: Federated Self-Evolution for Privacy-Constrained Multi-Environment LLM Agents Chen, Xiang Shi, Yuling Lan, Qizhen Qiu, Yuchao Wang, Min Gu, Xiaodong Yan, Yanfu Machine Learning Artificial Intelligence LLM agents are widely deployed in complex interactive tasks, yet privacy constraints often preclude centralized optimization and co-evolution across dynamic environments. Despite the demonstrated success of Federated Learning (FL) on static datasets, its effectiveness in open-ended, self-evolving agent systems remains largely unexplored. In such settings, the direct application of standard FL is particularly challenging, as heterogeneous tasks and sparse, trajectory-level reward signals give rise to severe gradient instability, which undermines the global optimization process. To bridge this gap, we propose Fed-SE, a Federated Self-Evolution framework for LLM agents that establishes a local evolution-global aggregation paradigm. Locally, agents employ parameter-efficient fine-tuning on filtered, high-return trajectories to achieve stable gradient updates. Globally, Fed-SE aggregates updates within a low-rank subspace, reducing communication cost across clients. Experiments across five heterogeneous environments demonstrate that Fed-SE improves average task success rates by 10\% over the state-of-the-art FedIT, validating its effectiveness in cross-environment knowledge transfer under privacy constraints. |
| title | Fed-SE: Federated Self-Evolution for Privacy-Constrained Multi-Environment LLM Agents |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2512.08870 |