Rahul.Kumar

I'm an AI Engineer transitioning into AI safety research. My work focuses on empirical evaluation of frontier model behavior under adversarial pressure, particularly how compliance-forcing instructions interact with epistemic boundaries in deployed systems. My research sits at the intersection of alignment evaluation, red-teaming methodology, and practical deployment safety.