CASE: Contrastive Activation for Class-Sensitive Explanations

arXiv:2506.07327 · cs.CV, cs.LG · Submitted 2025-06-08 · Read on arXiv

cs.CV, cs.LG

Submitted: 2025-06-08

Updated: 2026-09-18

Comments: 19 pages, 7 figures Accepted for publication in Springer Nature Machine Learning

DOI: 10.1007/s10994-026-07169-w

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Ablation — 18× in this paper · explained in 1 episode
• Fidelity — 14× in this paper · explained in 2 episodes
• Sparsity — 6× in this paper · explained in 4 episodes
• Consistency — 5× in this paper · explained in 3 episodes
• Interpretability — 5× in this paper · explained in 15 episodes
• Class Activation Mapping — 3× in this paper
• Ablation study — 2× in this paper · explained in 1 episode
• Aggregation — 2× in this paper · explained in 1 episode
• Benchmark — 2× in this paper · explained in 2 episodes
• Contrastive Activation for Saliency Estimation — 2× in this paper
• Redundancy — 2× in this paper · explained in 2 episodes
• Robustness — 2× in this paper · explained in 5 episodes
• Alignment — 1× in this paper · explained in 1 episode
• Clustering — 1× in this paper · explained in 2 episodes
• Computational overhead — 1× in this paper · explained in 2 episodes
• Convolutional Neural Networks (CNNs) — 1× in this paper · explained in 2 episodes
• Fine-Tuning — 1× in this paper · explained in 5 episodes
• Multimodal Reasoning — 1× in this paper · explained in 5 episodes
• Nadia — 1× in this paper · explained in 2 episodes
• Vision-language models — 1× in this paper · explained in 2 episodes
• Winter Conference on Applications of Computer Vision — 1× in this paper

Sources

• Network Dissection: Quantifying Interpretability of Deep Visual Representations
• Explaining Image Classifiers by Counterfactual Generation
• Latent Tree Models for Hierarchical Topic Detection
• Understanding Individual Decisions of CNNs via Contrastive Backpropagation
• Deep Residual Learning for Image Recognition
• Densely Connected Convolutional Networks
• CAMERAS: Enhanced Resolution And Sanity preserving Class Activation Mapping for image saliency
• HIVE: Evaluating the Human Interpretability of Visual Explanations
• The Disagreement Problem in Explainable Machine Learning: A Practitioner's Perspective
• An Evaluation of the Human-Interpretability of Explanation
• A ConvNet for the 2020s
• RISE: Randomized Input Sampling for Explanation of Black-box Models
• Evaluating the visualization of what a Deep Neural Network has learned
• Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization
• Score-CAM: Score-Weighted Visual Explanations for Convolutional Neural Networks
• On the (In)fidelity and Sensitivity for Explanations
• Learning Deep Features for Discriminative Localization

Related papers

• Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
• AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
• Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
• MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
• TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
• A Survey on Efficient Vision-Language-Action Models