Mochi: Aligning Pre-training and Inference for Efficient Graph Foundation Models via Meta-Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Mattos, João, Silva, Arlei
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909021094543360
author Mattos, João
Silva, Arlei
author_facet Mattos, João
Silva, Arlei
contents We propose Mochi, a Graph Foundation Model that addresses task unification and training efficiency by adopting a meta-learning based training framework. Prior models pre-train with reconstruction-based objectives such as link prediction, and assume that the resulting representations can be aligned with downstream tasks through a separate unification step such as class prototypes. We demonstrate through synthetic and real-world experiments that this procedure, while simple and intuitive, has limitations that directly affect downstream task performance. To address these limitations, Mochi pre-trains on few-shot episodes that mirror the downstream evaluation protocol, aligning the training objective with inference rather than relying on a post-hoc unification step. We show that Mochi, along with its more powerful variant Mochi++, achieves competitive or superior performance compared to existing Graph Foundation Models across 25 real-world graph datasets spanning node classification, link prediction, and graph classification, while requiring 8$\sim$27 times less training time than the strongest baseline.
format Preprint
id arxiv_https___arxiv_org_abs_2604_22031
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Mochi: Aligning Pre-training and Inference for Efficient Graph Foundation Models via Meta-Learning
Mattos, João
Silva, Arlei
Machine Learning
Artificial Intelligence
We propose Mochi, a Graph Foundation Model that addresses task unification and training efficiency by adopting a meta-learning based training framework. Prior models pre-train with reconstruction-based objectives such as link prediction, and assume that the resulting representations can be aligned with downstream tasks through a separate unification step such as class prototypes. We demonstrate through synthetic and real-world experiments that this procedure, while simple and intuitive, has limitations that directly affect downstream task performance. To address these limitations, Mochi pre-trains on few-shot episodes that mirror the downstream evaluation protocol, aligning the training objective with inference rather than relying on a post-hoc unification step. We show that Mochi, along with its more powerful variant Mochi++, achieves competitive or superior performance compared to existing Graph Foundation Models across 25 real-world graph datasets spanning node classification, link prediction, and graph classification, while requiring 8$\sim$27 times less training time than the strongest baseline.
title Mochi: Aligning Pre-training and Inference for Efficient Graph Foundation Models via Meta-Learning
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2604.22031