Evaluating the Instruction-Following Robustness of Large Language Models to Prompt Injection.

Zekun Li, Baolin Peng, Pengcheng He, Xifeng Yan

Browse the full EMNLP paper archive.