This work introduces a sampling-free, perturbation-based training framework based on continuous and differentiable masking that achieves competitive or superior attribution faithfulness compared to strong sampling-based baselines, while dramatically reducing computational cost and enabling substantially improved scalability.
Feature-robust Augmentation is introduced, which comprises diversified degradation-aware augmentation strategies, and a supervised contrastive learning pattern paired with a mean-teacher architecture that stabilizes features against augmentations through consistency constraints that wins the first place in ACM Multimed...
Zhu Xu, Jia-Qi Tang, Po-Kai Chen et al.· 0 citations
This work studies raw pixels, SD-VAE latents and DINOv2 as well as MAE representation-autoencoder features within a unified masked autoregressive rectified-flow model and shows that compression, reconstruction fidelity, token dimensionality, and visible semantic clustering do not individually predict generative behavio...
Marcel Plocher, B. Schölkopf, Andreas Geiger et al.· 0 citations
This paper proposes a parameter-free truncated pseudoinverse solver which removes collapsed directions prior to inversion, and achieves 81.42\% top-1 accuracy, outperforming prior post-training methods and fine-tuning-based baselines.
Quantitative evaluations demonstrate mask faithfulness, near-baseline classification performance across five classification tasks despite substantial masking of image information, and robustness to distribution shifts such as background swapping and natural adversarial examples.
Benjamin Formby, Kuang-Ching Wang, D. H. Smith· 0 citations
Understanding the reliability of model explanations remains a critical challenge in deep learning. Prior work has shown that saliency maps can be manipulated by optimizing the input using gradient-based methods, where gradients of the loss with respect to the input are computed to generate dense perturbations that alte...
Khoa Tran, Thai Huy Nguyen, Quan Minh Phan et al.· International Conference on...· 0 citations
Beyond efficiency and generalization, DAP natively provides increased robustness to adversarial perturbations and yields highly interpretable models, where the retained weights reliably encapsulate the most domain-invariant and task-critical representations.
