Papers › Towards Adversarial Robustness And Backdoor Mitigation in SSL

Towards Adversarial Robustness And Backdoor Mitigation in SSL

23 Mar 2024arXiv:2403.15918archive 2025-07-28

Aryan Satpathy, Nilaksh Singh, Dhruva Rajwade, Somesh Kumar

Self-Supervised Learning (SSL) has shown great promise in learning representations from unlabeled data. The power of learning representations without the need for human annotations has made SSL a widely used technique in real-world problems. However, SSL methods have recently been shown to be vulnerable to backdoor attacks, where the learned model can be exploited by adversaries to manipulate the learned representations, either through tampering the training data distribution, or via modifying the model itself. This work aims to address defending against backdoor attacks in SSL, where the adversary has access to a realistic fraction of the SSL training data, and no access to the model. We use novel methods that are computationally efficient as well as generalizable across different problem settings. We also investigate the adversarial robustness of SSL models when trained with our method, and show insights into increased robustness in SSL via frequency domain augmentations. We demonstrate the effectiveness of our method on a variety of SSL benchmarks, and show that our method is able to mitigate backdoor attacks while maintaining high performance on downstream tasks. Code for our work is available at github.com/Aryan-Satpathy/Backdoor

Paper PDF Code

Code

aryan-satpathy/backdoor officialmentioned in papermentioned on GitHubpytorchGPL-3.0 report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Adversarial Robustness Backdoor Attack Self-Supervised Learning

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

AdaGrad Attention BPE CTRL Dense Connections Dropout Gradient Clipping Layer Normalization Linear Layer Linear Warmup Multi-Head Attention ReLU Residual Connection Softmax

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections