dorsal/arxiv
View SchemaARM: Role-Conditioned Neuron Transplantation for Training-Free Generalist LLM Agent Merging
| Authors | Zhuoka Feng, Kang Chen, Sihan Zhao, Kai Xiong, Yaoning Wang, Minshen Yu, Junjie Nian, Changyi Xiao, Yixin Cao, Yugang Jiang |
|---|---|
| Categories | |
| ArXiv ID | 2601.07309vv1 |
| URL | https://arxiv.org/abs/2601.07309 |
| License | http://creativecommons.org/licenses/by/4.0/ |
Abstract
Interactive large language model agents have advanced rapidly, but most remain specialized to a single environment and fail to adapt robustly to other environments. Model merging offers a training-free alternative by integrating multiple experts into a single model. In this paper, we propose Agent-Role Merging (ARM), an activation-guided, role-conditioned neuron transplantation method for model merging in LLM agents. ARM improves existing merging methods from static natural language tasks to multi-turn agent scenarios, and over the generalization ability across various interactive environments. This is achieved with a well designed 3-step framework: 1) constructing merged backbones, 2) selection based on its role-conditioned activation analysis, and 3) neuron transplantation for fine-grained refinements. Without gradient-based optimization, ARM improves cross-benchmark generalization while enjoying efficiency. Across diverse domains, the model obtained via ARM merging outperforms prior model merging methods and domain-specific expert models, while demonstrating strong out-of-domain generalization.
{
"annotation_id": "a68d2381-9e2a-46ee-9736-d7b35f6a56a7",
"date_created": "2026-02-17T05:53:12.534000Z",
"date_modified": "2026-02-17T05:53:12.534000Z",
"file_hash": "b54efda95a6064f949555306db0625b86146158ef95c5cecc1030630d7d46eda",
"private": false,
"record": {
"abstract": "Interactive large language model agents have advanced rapidly, but most remain specialized to a single environment and fail to adapt robustly to other environments. Model merging offers a training-free alternative by integrating multiple experts into a single model. In this paper, we propose Agent-Role Merging (ARM), an activation-guided, role-conditioned neuron transplantation method for model merging in LLM agents. ARM improves existing merging methods from static natural language tasks to multi-turn agent scenarios, and over the generalization ability across various interactive environments. This is achieved with a well designed 3-step framework: 1) constructing merged backbones, 2) selection based on its role-conditioned activation analysis, and 3) neuron transplantation for fine-grained refinements. Without gradient-based optimization, ARM improves cross-benchmark generalization while enjoying efficiency. Across diverse domains, the model obtained via ARM merging outperforms prior model merging methods and domain-specific expert models, while demonstrating strong out-of-domain generalization.",
"arxiv_id": "2601.07309",
"authors": [
"Zhuoka Feng",
"Kang Chen",
"Sihan Zhao",
"Kai Xiong",
"Yaoning Wang",
"Minshen Yu",
"Junjie Nian",
"Changyi Xiao",
"Yixin Cao",
"Yugang Jiang"
],
"categories": [
"cs.AI",
"cs.LG"
],
"license": "http://creativecommons.org/licenses/by/4.0/",
"title": "ARM: Role-Conditioned Neuron Transplantation for Training-Free Generalist LLM Agent Merging",
"url": "https://arxiv.org/abs/2601.07309",
"version": "v1"
},
"schema_id": "dorsal/arxiv",
"source": {
"execution_id": "3fd3f2c9-a68f-4cd0-ac89-180508892c79",
"id": "arXiv Dataset",
"type": "Model",
"variant": "snapshot-2026-01-17",
"version": "0.1.0"
},
"user_id": 1000002
}