Zaloguj się
spostrzeżenie - MLLM Defense
AdaShield: Safeguarding Multimodal Large Language Models from Structure-based Attacks
AdaShield defends MLLMs against structure-based jailbreak attacks without fine-tuning or additional training.
AdaShield: Safeguarding Multimodal Large Language Models from Structure-based Attacks
Defending MLLMs against structure-based jailbreak attacks with AdaShield.
Produkty | Zasoby
© 2024 by Linnk AI