Two papers accepted at ICLR 2024! One paper on adversarial robustness in RL, the other paper is on imitating desired behaviors from multi-agent observations.