cs.LGNov 14, 2026

Learning Graph-Conditioned Predictors in Weight Space

Authors: Zhewei Chen, Hao Zhu, Ahad N. Zehmakan

Organizations: Australian National University, Australia · CSIRO Technology, Australia

Abstract

The same node attributes can imply different labels in different structural contexts. Yet fast graphless models typically treat topology as additional evidence supplied to a globally shared predictor. This assumes that structure augments the available information, rather than determining how node attributes should be interpreted. We introduce SCOPE (Structure-COnditioned PrEdictor), in which a small generator maps each node's cached structural fingerprint to coefficients over a shared low-rank basis, producing a context-specific update to a global weight matrix. The base predictor, basis, and generator remain shared across all nodes, while the resulting transformation adapts how attributes are processed in each relational environment. Once fingerprints are cached, inference requires only a forward pass and no online message passing. This design combines context-specific specialization with global parameter sharing, compact low-rank adaptation, and message-passing-free inference. On two large OGB node-classification graphs, conditioning the weights with raw positional signals yields clear gains over supplying the same signals through input concatenation. Further results under a structure-transductive held-out-node protocol and in teacher-free graph classification show that the mechanism extends beyond a particular distillation setting. These findings establish graph structure not only as predictive evidence, but as context that determines the predictor used to interpret that evidence.

Figures & tables

Appendix figures & tables11 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

Sep 29, 2026

Message Passing Does More with Less for In-Context Learning on Graphs

Achieving strong performance with graph neural networks (GNNs) typically requires training and hyperparameter tuning for each dataset, incurring repeated costs and effort. Graph in-context learning (ICL) avoids this by using a single pretrained model to predict unknown node labels directly from labeled context nodes. Existing approaches, however, rely on dense attention across nodes, making inference increasingly expensive as graphs grow. In this work, we present Ephris, a new graph in-context learner built on sparse message passing, scaling linearly with the number of node-feature entries and graph edges. Ephris is pretrained entirely on synthetic graphs generated from structural causal models with diverse graph structures and relational dynamics, exposing the model to varied dependencies among topology, features, and labels. We evaluate Ephris on 51 node-classification datasets against 15 extensively tuned GNNs and existing graph ICL methods under both high- and low-label train/validation/test splits. Across both settings, Ephris ranks first on all four aggregate measures: Elo, improvability, average rank, and accuracy. Its inference cost remains comparable to training a single GNN once, while being over 10 times faster than previous graph ICL models. Together, these results advance the performance-runtime Pareto frontier, demonstrating that strong graph ICL does not require dense attention. Code and model weights are available at https://github.com/nums-ai/ephris.
Oct 5, 2026

To Learn is to Wander: Learning Across Graphs and Tasks with Random Walks

Graph foundation models aim to transfer across graphs, feature spaces, relational schemas, and prediction tasks, yet existing approaches typically generalize only within particular graph modalities or tasks. We propose Wander, a graph foundation model designed to operate across these settings within a single pretrained checkpoint. Following the prior-predictive perspective, we formulate graph learning as completion of a partially observed graph. We realize this task-general view through a common interface based on random walks, allowing the same model to operate across homogeneous and multi-relational graphs with varying features, labels, and relational schemas. Wander can increase its structural context at inference time without changing its learned parameters and, under suitable assumptions, universally approximates the corresponding Bayes-optimal predictor on bounded connected graphs. Empirically, a single pretrained checkpoint achieves state-of-the-art or highly competitive results across node classification, homogeneous link prediction, and knowledge-graph link prediction. Moreover, joint pretraining across graph modalities and tasks preserves performance in specialized settings while enabling positive transfer and the composition of separately learned capabilities at inference time.
Sep 5, 2026

LoGIC: Budgeted Context Construction for Node-Level Graph In-Context Learning with Tabular Foundation Models

Tabular foundation models have become powerful graph learners. Systems such as G2T-FM and GraphPFN encode each node as a feature row and make predictions through in-context learning (ICL), with labeled rows serving as the prompt. Current protocols employ the complete training table as context, causing attention to scale quadratically with the labeled pool and introducing preprocessing and memory bottlenecks. We investigate context construction for node-level graph ICL: which labeled nodes and auxiliary unlabeled nodes should constitute the prompt for specified queries. We formulate this allocation in terms of two resources: a labeled-context budget for predictive evidence and an unlabeled-halo budget for adapter message passing without using label capacity. We present LoGIC, which retrieves labeled nodes via structural, feature-based, and coverage channels, shares each context across the queries in a graph-local cluster, incorporates an unlabeled halo for adapter backbones, and chooses the channel and context budget without test labels. Across three backbone configurations drawn from two model families on GraphLand, budgeted contexts maintain locally runnable full-context performance, stay competitive with published large-dataset results, and markedly lower peak memory requirements compared with full-context and whole-graph inference. They further permit frozen graph ICL on million-node graphs without retraining. Our analysis identifies when retrieval channels work best and connects their behavior with graph properties.