No article text was recorded for this version.
perceptually aligned gradients
Version 1 of 1206 papers282 conceptsfrozen 2026-10-04
Papers in this version
- Blog: Sparse Visual Counterfactual Explanations in Image Space
- Blog: Do Perceptually Aligned Gradients Imply Adversarial Robustness?
- Blog: MAGIC: Mask-Guided Image Synthesis by Inverting a Quasi-robust Classifier
- Blog: Adversarial Robustness as a Prior for Learned Representations
- Blog: Do Perceptually Aligned Gradients Imply Robustness?
- Blog: Do Perceptually Aligned Gradients Imply Robustness?
- Blog: Enhancing Consistency-Based Image Generation via Adversarialy-Trained Classification and Energy-Based Discrimination
- Blog: Stretching Beyond the Obvious: A Gradient-Free Framework to Unveil the Hidden Landscape of Visual Invariance
- Blog: Proper Network Interpretability Helps Adversarial Robustness in Classification
- Blog: Attack to Explain Deep Representation
- Blog: Understanding Invariance via Feedforward Inversion of Discriminatively Trained Classifiers
- Blog: Are Perceptually-Aligned Gradients a General Property of Robust Classifiers?
- Blog: L-WISE: Boosting Human Visual Category Learning Through Model-Based Image Selection and Enhancement
- Blog: Characterizing Model Robustness via Natural Input Gradients
- Blog: Is AI Interpretability Safe: the Relationship between Interpretability and Security of Machine Learning Models
- Blog: L-WISE: Boosting Human Image Category Learning Through Model-Based Image Selection And Enhancement
- Blog: Understanding Failures of Deep Networks via Robust Feature Extraction
- Blog: ZeroGrad : Mitigating and Explaining Catastrophic Overfitting in FGSM Adversarial Training
- Blog: Poisoned classifiers are not only backdoored, they are fundamentally broken
- Blog: Which Models have Perceptually-Aligned Gradients? An Explanation via Off-Manifold Robustness
- Blog: On the Benefits of Models with Perceptually-Aligned Gradients
- Blog: Improved and Interpretable Defense to Transferred Adversarial Examples by Jacobian Norm with Selective Input Gradient Regularization
- Blog: Generating Interpretable Counterfactual Explanations By Implicit Minimisation of Epistemic and Aleatoric Uncertainties
- Blog: CLIPAG: Towards Generator-Free Text-to-Image Generation
- Blog: Adversarial Robustness on In- and Out-Distribution Improves Explainability
- Blog: SAM: The Sensitivity of Attribution Methods to Hyperparameters
- Blog: Salient ImageNet: How to discover spurious features in Deep Learning?
- Blog: Image Synthesis with a Single (Robust) Classifier
- Blog: Generating Interpretable Counterfactual Explanations By Implicit Minimisation of Epistemic and Aleatoric Uncertainties
- Blog: Towards Adversarially Robust Knowledge Graph Embeddings
- Blog: Visual explanations for the detection of diabetic retinopathy from retinal fundus images
- Blog: Diffusion Visual Counterfactual Explanations
- Blog: Text-to-Image Generation Via Energy-Based CLIP
- Blog: DiffAug: A Diffuse-and-Denoise Augmentation for Training Robust Classifiers
- Blog: Adversarial training improves model interpretability in single-cell RNA-seq analysis
- Blog: Understanding Failures of Deep Networks via Robust Feature Extraction
- Blog: Robustness May Be at Odds with Accuracy
- Blog: Causal ImageNet: How to discover spurious features in Deep Learning?
- Blog: Does Interpretability of Neural Networks Imply Adversarial Robustness?
- Blog: Inverting Adversarially Robust Networks for Image Synthesis
- Blog: Generating Realistic Counterfactuals for Retinal Fundus and OCT Images using Diffusion Models
- Blog: Robust Explanation for Free or At the Cost of Faithfulness
- Blog: Spurious Features Everywhere - Large-Scale Detection of Harmful Spurious Features in ImageNet
- Blog: Concise Explanations of Neural Networks using Adversarial Training
- Blog: Jacobian Adversarially Regularized Networks for Robustness
- Blog: Exploring Conceptual Soundness with TruLens
- Blog: Towards Interpretable Deep Neural Networks by Leveraging Adversarial Examples
- Blog: On the Benefits of Attributional Robustness
- Blog: BIGRoC: Boosting Image Generation via a Robust Classifier
- Blog: Fine-Tuning Text-To-Image Diffusion Models for Class-Wise Spurious Feature Generation
- Blog: Training Deep Neural Networks for Interpretability and Adversarial Robustness
- Blog: Fast Training of Deep Neural Networks Robust to Adversarial Perturbations
- Blog: Do Input Gradients Highlight Discriminative Features?
- Blog: Stretching Beyond the Obvious: A Gradient-Free Framework to Unveil the Hidden Landscape of Visual Invariance
- Blog: Explainable Machine Learning in Deployment
- Blog: Scaling Adversarial Training to Large Perturbation Bounds
- Blog: On the Connection Between Adversarial Robustness and Saliency Map Interpretability
- Blog: Feature Prioritization and Regularization Improve Standard Accuracy and Adversarial Robustness
- Blog: Exploring the Connection between Robust and Generative Models
- Blog: Enhancing Diffusion-Based Image Synthesis with Robust Classifier Guidance
Showing 60 of 206 papers.
History & citation
Current · Permanent link
- Perceptually Aligned Gradients in Robust Neural Networks merged into this direction (75% overlap)
- Aligning Neural Representations Through Adversarial Robustness merged into this direction (85% overlap)
Cite this direction
Latest version (a living page):
@misc{lacuna_perceptually_aligned_gradients_d074f0c8,
title = {perceptually aligned gradients},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/perceptually-aligned-gradients/rdr_d074f0c8cf6c424ba3c64f65ea8c891d},
note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-20},
}This exact version (version 1, frozen text):
@misc{lacuna_perceptually_aligned_gradients_d074f0c8_v1,
title = {perceptually aligned gradients},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/perceptually-aligned-gradients/rdr_d074f0c8cf6c424ba3c64f65ea8c891d/v/1},
note = {Version 1, frozen 2026-10-04. Accessed 2026-11-20},
}