1 paper · 2 filters
Zeru Shi, Zhenting Wang, Yongye Su +5
While automatic prompt generation methods have recently received significant attention, their robustness remains poorly understood. In this paper, we introduce PertBench, a compreh…