When ‘Competency’ in Reasoning Opens the Door to Vulnerability: Jailbreaking LLMs via Novel Complex Ciphers
Published in NeurIPS 2025 Reliable ML from Unreliable Data Workshop (non-archival), 2024
Recommended citation: D. Handa, Z. Zhang, A. Saeidi, S. Kumbhar, and C. Baral. "When Competency in Reasoning Opens the Door to Vulnerability: Jailbreaking LLMs via Novel Complex Ciphers." NeurIPS 2025 Reliable ML from Unreliable Data Workshop.
Download Paper