What Makes VLMs Robust? Towards Reconciling Robustness and Accuracy in Vision-Language Models
Project website: https://summu77.github.io/R-Adapt/
Paper: https://arxiv.org/abs/2603.12799
The code for R-Adapt is coming soon.
We will release the training and evaluation code, together with related resources, in this repository.