TY - RPRT TI - LLMs are Vulnerable to Malicious Prompts Disguised as Scientific Language AU - Yubin Ge AU - Neeraja Kirtane AU - Hao Peng AU - Dilek Hakkani-Tür PY - 2025 UR - https://arxiv.org/abs/2501.14073 ID - 2501.14073 ER -