タグ #敵対的攻撃 に関するスライド

slide-thumbnail

【DL輪読会】Backtracking Improves Generation Safety
user-img Deep Learning JP 1K
slide-thumbnail

【DL輪読会】AdvWeb: Controllable Black-box Attacks on VLM-powered Web Agents
user-img Deep Learning JP 349
slide-thumbnail

【DL輪読会】COLD-Attack: Jailbreaking LLMs with Stealthiness and Controllability (ICML2024)
user-img Deep Learning JP 1.5K
#敵対的攻撃
#LLM