mlsanityData to 7 Oct 2026

Predicting the generalization gap in neural networks using topological data analysis

arXiv 2203.12330 ↗First posted 23 Mar 2022cs.LG math.ATcited by 9

Not on the map. Why?

Abstract (arXiv)

Understanding how neural networks generalize on unseen data is crucial for designing more robust and reliable models. In this paper, we study the generalization gap of neural networks using methods from topological data analysis. For this purpose, we compute homological persistence diagrams of weighted graphs constructed from neuron activation correlations after a training phase, aiming to capture patterns that are linked to the generalization capacity of the network. We compare the usefulness of different numerical summaries from persistence diagrams and show that a combination of some of them can accurately predict and partially explain the generalization gap without the need of a test set. Evaluation on two computer vision recognition tasks (CIFAR10 and SVHN) shows competitive generalization gap prediction when compared against state-of-the-art methods.

Key facts. Cited by 9 papers in the map (data through 7 Oct 2026). Builds on 39 of its 77 references that are in the map. First posted on arXiv 23 Mar 2022.

Builds on

39 of 77 references.

References · 39

  1. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
  2. 2014Explaining and Harnessing Adversarial Examplescited by 7,287
  3. 2017Spectrally-normalized margin bounds for neural networkscited by 681
  4. 2017Exploring Generalization in Deep Learningcited by 595
  5. 2015Train faster, generalize better: Stability of stochastic gradient descentcited by 547
  6. 2017Computing Nonvacuous Generalization Bounds for Deep (Stochastic) Neural Networks with Many More Parameters than Training Datacited by 527
  7. 2019Fantastic Generalization Measures and Where to Find Themcited by 507
  8. 2017Size-Independent Sample Complexity of Neural Networkscited by 289
  9. 2017The Marginal Value of Adaptive Gradient Methods in Machine Learningcited by 248
  10. 2018Why do deep convolutional networks generalize so poorly to small image transformations?cited by 198
  11. 2017Generalization in Deep Learningcited by 146
  12. 2017A Bayesian Perspective on Generalization and Stochastic Gradient Descentcited by 134
27 more
  1. 2018Predicting the Generalization Gap in Deep Networks with Margin Distributionscited by 111
  2. 2017Stochastic gradient descent performs variational inference, converges to limit cycles for deep networkscited by 108
  3. 2017Fisher-Rao Metric, Geometry, and Complexity of Neural Networkscited by 99
  4. 2018Neural Persistence: A Complexity Measure for Deep Neural Networks Using Algebraic Topologycited by 73
  5. 2021Intrinsic Dimension, Persistent Homology and Generalization in Neural Networkscited by 70
  6. 2020Hausdorff Dimension, Heavy Tails, and Generalization in Neural Networkscited by 62
  7. 2018Geometry Score: A Method For Comparing Generative Adversarial Networkscited by 56
  8. 2020giotto-tda: A Topological Data Analysis Toolkit for Machine Learning and Data Explorationcited by 55
  9. 2020Computing the Testing Error without a Testing Setcited by 54
  10. 2019Generalization in Deep Networks: The Role of Distance from Initializationcited by 49
  11. 2020NeurIPS 2020 Competition: Predicting Generalization in Deep Learningcited by 41
  12. 2019Adaptive Hierarchical Down-Sampling for Point Cloud Classificationcited by 40
  13. 2018On Characterizing the Capacity of Neural Networks using Algebraic Topologycited by 37
  14. 2018Topological Approaches to Deep Learningcited by 29
  15. 2019A Framework for Differential Calculus on Persistence Barcodescited by 29
  16. 2020Representation Based Complexity Measures for Predicting Generalization in Deep Learningcited by 29
  17. 2018Exposition and Interpretation of the Topology of Neural Networkscited by 20
  18. 2020Evaluating the Disentanglement of Deep Generative Models through Manifold Topologycited by 20
  19. 2018Topological Data Analysis of Decision Boundaries with Application to Model Selectioncited by 19
  20. 2018Persistent Homology and the Upper Box Dimensioncited by 15
  21. 2021Predicting Deep Neural Network Generalization with Perturbation Response Curvescited by 14
  22. 2021Robustness to Augmentations as a Generalization metriccited by 13
  23. 2020Finding the Homology of Decision Boundaries with Active Learningcited by 12
  24. 2021giotto-ph: A Python Library for High-Performance Computation of Persistent Homology of Vietoris-Rips Filtrationscited by 10
  25. 2015Comparing persistence diagrams through complex vectorscited by 9
  26. 2019Path homologies of deep feedforward networkscited by 6
  27. 2020Ranking Deep Learning Generalization using Label Variation in Latent Geometry Graphscited by 6

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 2; 2024: 0; 2025: 3; 2026: 4.

9 papers.

  1. 2023Topological Data Analysis for Neural Network Analysis: A Comprehensive Surveycited by 9
  2. 2023Importance attribution in neural networks by means of persistence landscapes of time seriescited by 1
  3. 2025One Period to Rule Them All: Identifying Critical Learning Periods in Deep Networkscited by 1
  4. 2025HOLE: Homological Observation of Latent Embeddings for Neural Network Interpretabilitycited by 1
  5. 2026TopoGeoScore: A Self-Supervised Source-Only Geometric Framework for OOD Checkpoint Selectioncited by 1
  6. 2026Deciphering Two Training Clocks in Grokking via Deep Linear Network Theory with Conditional ReLU Reductioncited by 1
  7. 2025Is Supervised Learning Really That Different from Unsupervised?cited by 0
  8. 2026Upper Bounds on the Generalization Error of Deep Learning Models via Local Robustness and Stabilitycited by 0
  9. 2026BettiSplit: Topology-Guided Privacy-Aware Split Learning Against Feature Inversion and Gradient Leakagecited by 0