mlsanityData to 7 Oct 2026

Geometric Remove-and-Retrain (GOAR): Coordinate-Invariant eXplainable AI Assessment

arXiv 2407.12401 ↗First posted 17 Jul 2024cs.LG cs.CVcited by 3

Not on the map. Why?

Abstract (arXiv)

Identifying the relevant input features that have a critical influence on the output results is indispensable for the development of explainable artificial intelligence (XAI). Remove-and-Retrain (ROAR) is a widely accepted approach for assessing the importance of individual pixels by measuring changes in accuracy following their removal and subsequent retraining of the modified dataset. However, we uncover notable limitations in pixel-perturbation strategies. When viewed from a geometric perspective, we discover that these metrics fail to discriminate between differences among feature attribution methods, thereby compromising the reliability of the evaluation. To address this challenge, we introduce an alternative feature-perturbation approach named Geometric Remove-and-Retrain (GOAR). Through a series of experiments with both synthetic and real datasets, we substantiate that GOAR transcends the limitations of pixel-centric metrics.

Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Builds on 34 of its 42 references that are in the map. First posted on arXiv 17 Jul 2024.

Builds on

34 of 42 references.

References · 34

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  3. 2020Denoising Diffusion Implicit Modelscited by 9,655
  4. 2016Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localizationcited by 6,690
  5. 2017A Unified Approach to Interpreting Model Predictionscited by 6,488
  6. 2016"Why Should I Trust You?": Explaining the Predictions of Any Classifiercited by 5,040
  7. 2017Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithmscited by 4,283
  8. 2017Axiomatic Attribution for Deep Networkscited by 3,350
  9. 2013Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Mapscited by 2,348
  10. 2017Learning Important Features Through Propagating Activation Differencescited by 1,454
  11. 2022Diffusion Posterior Sampling for General Noisy Inverse Problemscited by 1,375
  12. 2017SmoothGrad: removing noise by adding noisecited by 1,097
22 more
  1. 2014Striving for Simplicity: The All Convolutional Netcited by 983
  2. 2021ILVR: Conditioning Method for Denoising Diffusion Probabilistic Modelscited by 634
  3. 2020Captum: A unified and generic model interpretability library for PyTorchcited by 542
  4. 2022Improving Diffusion Models for Inverse Problems using Manifold Constraintscited by 501
  5. 2018A Benchmark for Interpretability Methods in Deep Neural Networkscited by 467
  6. 2019On the (In)fidelity and Sensitivity for Explanationscited by 309
  7. 2016Not Just a Black Box: Learning Important Features Through Propagating Activation Differencescited by 295
  8. 2020Evaluating Explainable AI: Which Algorithmic Explanations Help Users Predict Model Behavior?cited by 189
  9. 2020Evaluating and Aggregating Feature-based Model Explanationscited by 185
  10. 2020Debugging Tests for Model Explanationscited by 146
  11. 2022OpenXAI: Towards a Transparent Evaluation of Model Explanationscited by 125
  12. 2019Sanity Checks for Saliency Metricscited by 119
  13. 2022A Consistent and Efficient Evaluation Strategy for Attribution Methodscited by 100
  14. 2020On quantitative aspects of model interpretabilitycited by 71
  15. 2021Have We Learned to Explain?: How Interpretability Methods Can Learn to Encode Predictions in their Interpretationscited by 59
  16. 2021Synthetic Benchmarks for Scientific Research in Explainable Machine Learningcited by 59
  17. 2020An Explainable Machine Learning Model for Early Detection of Parkinson's Disease using LIME on DaTscan Imagerycited by 25
  18. 2023Which Models have Perceptually-Aligned Gradients? An Explanation via Off-Manifold Robustnesscited by 20
  19. 2020Self-explaining AI as an alternative to interpretable AIcited by 18
  20. 2022The Manifold Hypothesis for Gradient-Based Explanationscited by 15
  21. 2022Towards Good Practices in Evaluating Transfer Adversarial Attackscited by 14
  22. 2023On Pitfalls of textitRemOve-And-Retrain: Data Processing Inequality Perspectivecited by 2

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 1; 2026: 2.

3 papers.

  1. 2025Rethinking Explanation Evaluation under the Retraining Schemecited by 0
  2. 2026Missingness Bias Calibration in Feature Attribution Explanationscited by 0
  3. 2026AIM: Adversarial Information Masking for Faithfulness Evaluation of Saliency Mapscited by 0