Out-of-Distribution Graph Models Merging

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wang, Yidi, Qiao, Ziyue, Gu, Jiawei, Zheng, Xubin, Wang, Pengyang, Pei, Xiaobing, Luo, Xiao
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912969881812992
author Wang, Yidi
Qiao, Ziyue
Gu, Jiawei
Zheng, Xubin
Wang, Pengyang
Pei, Xiaobing
Luo, Xiao
author_facet Wang, Yidi
Qiao, Ziyue
Gu, Jiawei
Zheng, Xubin
Wang, Pengyang
Pei, Xiaobing
Luo, Xiao
contents This paper studies a novel problem of out-of-distribution graph models merging, which aims to construct a generalized model from multiple graph models pre-trained on different domains with distribution discrepancy. This problem is challenging because of the difficulty in learning domain-invariant knowledge implicitly in model parameters and consolidating expertise from potentially heterogeneous GNN backbones. In this work, we propose a graph generation strategy that instantiates the mixture distribution of multiple domains. Then, we merge and fine-tune the pre-trained graph models via a MoE module and a masking mechanism for generalized adaptation. Our framework is architecture-agnostic and can operate without any source/target domain data. Both theoretical analysis and experimental results demonstrate the effectiveness of our approach in addressing the model generalization problem.
format Preprint
id arxiv_https___arxiv_org_abs_2506_03674
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Out-of-Distribution Graph Models Merging
Wang, Yidi
Qiao, Ziyue
Gu, Jiawei
Zheng, Xubin
Wang, Pengyang
Pei, Xiaobing
Luo, Xiao
Machine Learning
This paper studies a novel problem of out-of-distribution graph models merging, which aims to construct a generalized model from multiple graph models pre-trained on different domains with distribution discrepancy. This problem is challenging because of the difficulty in learning domain-invariant knowledge implicitly in model parameters and consolidating expertise from potentially heterogeneous GNN backbones. In this work, we propose a graph generation strategy that instantiates the mixture distribution of multiple domains. Then, we merge and fine-tune the pre-trained graph models via a MoE module and a masking mechanism for generalized adaptation. Our framework is architecture-agnostic and can operate without any source/target domain data. Both theoretical analysis and experimental results demonstrate the effectiveness of our approach in addressing the model generalization problem.
title Out-of-Distribution Graph Models Merging
topic Machine Learning
url https://arxiv.org/abs/2506.03674