mlsanityData to 7 Oct 2026

Are Graph Neural Networks Miscalibrated?

arXiv 1905.02296 ↗First posted 7 May 2019cs.LG cs.CV stat.MLcited by 14

Not on the map. Why?

Abstract (arXiv)

Graph Neural Networks (GNNs) have proven to be successful in many classification tasks, outperforming previous state-of-the-art methods in terms of accuracy. However, accuracy alone is not enough for high-stakes decision making. Decision makers want to know the likelihood that a specific GNN prediction is correct. For this purpose, obtaining calibrated models is essential. In this work, we perform an empirical evaluation of the calibration of state-of-the-art GNNs on multiple datasets. Our experiments show that GNNs can be calibrated in some datasets but also badly miscalibrated in others, and that state-of-the-art calibration methods are helpful but do not fix the problem.

Key facts. Cited by 14 papers in the map (data through 7 Oct 2026). First posted on arXiv 7 May 2019.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 3; 2022: 3; 2023: 2; 2024: 3; 2025: 2; 2026: 1.

14 papers.

Sort:

  1. 2021Be Confident! Towards Trustworthy Graph Neural Networks via Confidence Calibrationcited by 91
  2. 2022What Makes Graph Neural Networks Miscalibrated?cited by 39
  3. 2023SimCalib: Graph Neural Network Calibration based on Similarity between Nodescited by 10
  4. 2022On Calibration of Graph Neural Networks for Node Classificationcited by 9
  5. 2022A Graph Is More Than Its Nodes: Towards Structured Uncertainty-Aware Learning on Graphscited by 6
  6. 2023On the Temperature of Bayesian Graph Neural Networks for Conformal Predictioncited by 4
  7. 2024In-n-Out: Calibrating Graph Neural Networks for Link Predictioncited by 3
  8. 2024Uncertainty Quantification on Graph Learning: A Surveycited by 3
  9. 2024GETS: Ensemble Temperature Scaling for Calibration in Graph Neural Networkscited by 3
  10. 2025Multimodal Graph Representation Learning for Robust Surgical Workflow Recognition with Adversarial Feature Disentanglementcited by 3
  11. 2021CoG: a Two-View Co-training Framework for Defending Adversarial Attacks on Graphcited by 1
  12. 2025Efficient quantification on large-scale networkscited by 1
2 more
  1. 2021Set Twister for Single-hop Node Classificationcited by 0
  2. 2026HoTS: Homophily-Aware Temperature Scaling for Graph Neural Network Calibrationcited by 0