作者位次 #7/12SIGIR 2026Poster· Information Retrieval (Resource)
EVADE-Bench: Multimodal Benchmark for Evaluating and Enhancing Evasive Content Detection
Ancheng Xu · Zhihao Yang · Jingpeng Li · Guanghu Yuan · Longze Chen · Liang Yan · Jiehui Zhou · Zhen Qin · Hengyu Chang · Yukun Chen · Hamid Alinejad-Rokny · Min Yang
摘要
E-commerce platforms increasingly rely on Large Language Models (LLMs) and Vision Language Models (VLMs) to detect illicit or misleading product content. However, these models remain vulnerable to evasive content, which refers to inputs that have been deliberately modified through techniques such...