Neel Kant

Papers

Total Citations

H-Index

About

Neel Kant is a researcher whose work sits at the intersection of artificial intelligence and game theory, with a particular focus on the robustness and security of deep reinforcement learning systems. His most influential contribution is the seminal paper "Adversarial Policies: Attacking Deep Reinforcement Learning" (2019), which has garnered 92 citations and fundamentally changed how the field thinks about multi-agent safety. In this work, Kant demonstrated that an adversarial agent can learn a policy that, without directly modifying another agent's observations, reliably induces that agent to fail—a finding with profound implications for deploying RL in high-stakes environments like autonomous driving or robotics. Beyond this flagship study, Kant’s research explores how agents can learn to exploit vulnerabilities in other learned systems, effectively bridging the gap between adversarial machine learning and multi-agent dynamics. His work is notable not only for its technical rigor but also for its practical urgency: by revealing that even state-of-the-art RL policies are fragile under coordinated attack, Kant has helped catalyze a new line of research into robust and certifiably safe multi-agent learning. For students and researchers, his contributions serve as both a cautionary tale and a call to action for building more resilient AI.