Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Graph Transductive Sharpening: Leveraging Unlabeled Predictions in Node Classification

About

In the transductive setting, where the full graph is observed but node labels are only partially available, progress in semi-supervised node classification has largely focused on architectural innovation. In this paper, we revisit an orthogonal axis: the training objective. We start from a simple observation: transductive models produce predictions for every node during training, including nodes without labels. These unlabeled-node predictions may contain useful training signal, but standard supervised objectives discard them because no ground-truth labels are available. Inspired by the decomposition of cross-entropy into a label-dependent alignment term and a label-independent entropy term, we propose prediction confidence as a natural way to extract this signal in the absence of labels. This motivates Transductive Sharpening (TS): a loss-level modification that minimizes prediction entropy on unlabeled nodes while counterbalancing this effect on labeled nodes. We evaluate Transductive Sharpening across a wide range of node-classification benchmarks and observe consistent performance improvements without requiring any changes to the backbone architecture. Code is available at https://github.com/transductive-sharpening/tunedGNN.

Brown Zaz, Mar Gonz\`alez I Catal\`a, Ferran Hernandez Caralt, Moshe Eliasof, Pietro Li\`o• 2026

Related benchmarks

TaskDatasetResultRank
Node ClassificationChameleon
Accuracy45.3
867
Node ClassificationSquirrel
Accuracy44.57
786
Node ClassificationPubmed
Accuracy80.74
627
Node ClassificationCora
Accuracy85.74
583
Node Classificationogbn-arxiv (test)
Accuracy73.3
497
Node ClassificationRoman-Empire
Accuracy91.66
327
Node Classificationamazon-ratings
Accuracy56.72
309
Node ClassificationCS
Accuracy96.29
175
Node ClassificationComputer
Accuracy94.12
159
Node ClassificationPhoto
Accuracy96.69
153
Showing 10 of 12 rows

Other info

Follow for update