Hydra: Pruning adversarially robust neural networks

V Sehwag, S Wang, P Mittal… - Advances in Neural …, 2020 - proceedings.neurips.cc
In safety-critical but computationally resource-constrained applications, deep learning faces
two key challenges: lack of robustness against adversarial attacks and large neural network …

Advclip: Downstream-agnostic adversarial examples in multimodal contrastive learning

Z Zhou, S Hu, M Li, H Zhang, Y Zhang… - Proceedings of the 31st …, 2023 - dl.acm.org
Multimodal contrastive learning aims to train a general-purpose feature extractor, such as
CLIP, on vast amounts of raw, unlabeled paired image-text data. This can greatly benefit …

Bridging mode connectivity in loss landscapes and adversarial robustness

P Zhao, PY Chen, P Das, KN Ramamurthy… - ar** a robust and constructive framework for
tackling complex learning tasks. Consequently, it is widely utilized in many security-critical …

Sparsity winning twice: Better robust generalization from more efficient training

T Chen, Z Zhang, P Wang, S Balachandra… - arxiv preprint arxiv …, 2022 - arxiv.org
Recent studies demonstrate that deep networks, even robustified by the state-of-the-art
adversarial training (AT), still suffer from large robust generalization gaps, in addition to the …