Entrar
insight - Transfer Attacks Defense Strategy
Defending Against Transfer Attacks from Public Models at ICLR 2024
Proposing a practical defense strategy against transfer attacks from public models, PUBDEF outperforms white-box adversarial training with minimal loss in clean accuracy.
ICLR 2024: PUBDEF - Defending Against Transfer Attacks from Public Models
Adversarial attacks are a significant threat, but defending against transfer attacks from public models is crucial for security-sensitive applications.
Produtos | Recursos
© 2024 by Linnk AI