mlsanityData to 14 Oct 2026

On the Prediction Instability of Graph Neural Networks

arXiv 2205.10070 ↗First posted 20 May 2022cs.LGcited by 6

Not on the map. Why?

Abstract (arXiv)

Instability of trained models, i.e., the dependence of individual node predictions on random factors, can affect reproducibility, reliability, and trust in machine learning systems. In this paper, we systematically assess the prediction instability of node classification with state-of-the-art Graph Neural Networks (GNNs). With our experiments, we establish that multiple instantiations of popular GNN models trained on the same data with the same model hyperparameters result in almost identical aggregated performance but display substantial disagreement in the predictions for individual nodes. We find that up to one third of the incorrectly classified nodes differ across algorithm runs. We identify correlations between hyperparameters, node properties, and the size of the training set with the stability of predictions. In general, maximizing model performance implicitly also reduces model instability.

Key facts. Cited by 6 papers in the map (data through 14 Oct 2026). Builds on 19 of its 24 references that are in the map. First posted on arXiv 20 May 2022.

Builds on

19 of 24 references.

References · 19

  1. 2016Semi-Supervised Classification with Graph Convolutional Networkscited by 11,361
  2. 2017Graph Attention Networkscited by 8,923
  3. 2019Fast Graph Representation Learning with PyTorch Geometriccited by 2,434
  4. 2020Open Graph Benchmark: Datasets for Machine Learning on Graphscited by 2,163
  5. 2019Similarity of Neural Network Representations Revisitedcited by 1,755
  6. 2018Pitfalls of Graph Neural Network Evaluationcited by 1,007
  7. 2016Revisiting Semi-Supervised Learning with Graph Embeddingscited by 976
  8. 2020Benchmarking Graph Neural Networkscited by 680
  9. 2018Adversarial Attacks on Neural Networks for Graph Datacited by 397
  10. 2020Wiki-CS: A Wikipedia-Based Benchmark for Graph Neural Networkscited by 260
  11. 2019Certifiable Robustness and Robust Training for Graph Convolutional Networkscited by 59
  12. 2021Randomness In Neural Network Training: Characterizing The Impact of Toolingcited by 59
7 more
  1. 2021Nondeterminism and Instability in Neural Network Optimizationcited by 48
  2. 2021On the Reproducibility of Neural Network Predictionscited by 29
  3. 2021Stability of Graph Convolutional Neural Networks to Stochastic Perturbationscited by 18
  4. 2021Churn Reduction via Distillationcited by 17
  5. 2020The Effects of Randomness on the Stability of Node Embeddingscited by 15
  6. 2020Anti-Distillation: Improving reproducibility of deep networkscited by 13
  7. 2022Model Stability with Continuous Data Updatescited by 9

Built on by

Cited by year (2026 to 14 Oct)

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

6 papers.

  1. 2023Similarity of Neural Network Models: A Survey of Functional and Representational Measurescited by 131
  2. 2023Geometric instability of graph neural networks on large graphscited by 2
  3. 2023Distilling Influences to Mitigate Prediction Churn in Graph Neural Networkscited by 0
  4. 2024Exploring End-to-end Differentiable Neural Charged Particle Tracking -- A Loss Landscape Perspectivecited by 0
  5. 2025Revisiting the Relation Between Robustness and Universalitycited by 0
  6. 2026The Impact of Dimensionality on the Stability of Node Embeddingscited by 0