作者位次 #5/5NeurIPS 2026Poster· Computer Vision / Vision Models & Multimodal
What to Perturb, How to Propagate: A Graph-Guided Transferable Attack on VLP Models
Yongjae Kim · Sungbin Park · Hoon Ji · Sukmin Yun · Yeonjoon Lee
首次获取该候选人的完整档案及 StartLux 评估共 20 积分；每位用户单独解锁，之后查看免费。主动刷新 20 积分。
实时查找电子邮箱，费用不超过20积分；未找到邮箱不扣费。
Yongjae Kim · Sungbin Park · Hoon Ji · Sukmin Yun · Yeonjoon Lee
Jungmin Lee · Peizhuo Lv · Yeonjoon Lee
As Large Language Model applications (LLM apps) become widespread, system prompts that determine app behavior are increasingly regarded as intellectual property, raising concerns about leakage. Recent studies show that this threat is no longer theoretical, revealing the prevalence of cloned apps...