mlsanityData to 7 Oct 2026

The Confusion is Real: GRAPHIC -- A Network Science Approach to Confusion Matrices in Deep Learning

arXiv 2602.19770 ↗First posted 23 Feb 2026cs.LG cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Explainable artificial intelligence has emerged as a promising field of research to address reliability concerns in artificial intelligence. Despite significant progress in explainable artificial intelligence, few methods provide a systematic way to visualize and understand how classes are confused and how their relationships evolve as training progresses. In this work, we present GRAPHIC, an architecture-agnostic approach that analyzes neural networks on a class level. It leverages confusion matrices derived from intermediate layers using linear classifiers. We interpret these as adjacency matrices of directed graphs, allowing tools from network science to visualize and quantify learning dynamics across training epochs and intermediate layers. GRAPHIC provides insights into linear class separability, dataset issues, and architectural behavior, revealing, for example, similarities between flatfish and man and labeling ambiguities validated in a human study. In summary, by uncovering real confusions, GRAPHIC offers new perspectives on how neural networks learn. The code is available at https://github.com/Johanna-S-Froehlich/GRAPHIC.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 31 of its 69 references that are in the map. First posted on arXiv 23 Feb 2026.

Builds on

31 of 69 references.

References · 31

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2015Distilling the Knowledge in a Neural Networkcited by 10,076
  3. 2016Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localizationcited by 6,690
  4. 2016"Why Should I Trust You?": Explaining the Predictions of Any Classifiercited by 5,040
  5. 2019Similarity of Neural Network Representations Revisitedcited by 1,695
  6. 2017Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)cited by 1,309
  7. 2016Understanding intermediate layers using linear classifier probescited by 1,279
  8. 2020Concept Bottleneck Modelscited by 1,022
  9. 2021Do Vision Transformers See Like Convolutional Neural Networks?cited by 671
  10. 2017BranchyNet: Fast Inference via Early Exiting from Deep Neural Networkscited by 591
  11. 2021Pervasive Label Errors in Test Sets Destabilize Machine Learning Benchmarkscited by 388
  12. 2019Deep Multi-modal Object Detection and Semantic Segmentation for Autonomous Driving: Datasets, Methods, and Challengescited by 313
19 more
  1. 2020A Survey on Neural Network Interpretabilitycited by 291
  2. 2022Post-hoc Concept Bottleneck Modelscited by 289
  3. 2019Intrinsic dimension of data representations in deep neural networkscited by 276
  4. 2021Learning with Noisy Labels Revisited: A Study Using Real-World Human Annotationscited by 253
  5. 2015Interpretable classifiers using rules and Bayesian analysis: Building a better stroke prediction modelcited by 140
  6. 2023The geometry of hidden representations of large transformer modelscited by 108
  7. 2016Visualizing Large-scale and High-dimensional Datacited by 79
  8. 2017Do Convolutional Neural Networks Learn Class Hierarchy?cited by 61
  9. 2022Concept Activation Regions: A Generalized Framework For Concept-Based Explanationscited by 57
  10. 2024Explainable Artificial Intelligence: A Survey of Needs, Techniques, Applications, and Future Directioncited by 46
  11. 2022Optimizing Relevance Maps of Vision Transformers Improves Robustnesscited by 41
  12. 2019Characterizing the Decision Boundary of Deep Neural Networkscited by 32
  13. 2018t-SNE-CUDA: GPU-Accelerated t-SNE and its Applications to Modern Datacited by 19
  14. 2022Unsupervised visualization of image datasets using contrastive learningcited by 16
  15. 2019ConfusionFlow: A model-agnostic visualization for temporal analysis of classifier confusioncited by 15
  16. 2025Explainable artificial intelligence (XAI): from inherent explainability to large language modelscited by 12
  17. 2022Training a Vision Transformer from scratch in less than 24 hours with 1 GPUcited by 6
  18. 2025Demystifying the Accuracy-Interpretability Trade-Off: A Case Study of Inferring Ratings from Reviewscited by 5
  19. 2022On the inadequacy of nominal assortativity for assessing homophily in networkscited by 3

Built on by

Not cited yet.