1 paper · 1 filter
Shuze Liu, Kaixiang Zhao, Runyang Xu +4
Large language models (LLMs) deployed through text-only APIs face model extraction risks, as adversaries can collect their responses to train surrogates that reproduce their capabi…