Wen Xing

Can Rea­son­ing Models Obfus­cate Rea­son­ing? Stress-Test­ing Chain-of-Thought Monitorability

Vuln­er­a­bil­ity in Trusted Mon­i­tor­ing and Mitigations