End-to-end QP-based policies: A unified perspective on robust control and robot learning
This paper establishes connections to existing policy representations and control paradigms, including robust control, multilayer perceptrons, and robot learning, and interprets end-to-end QP policies as a common abstraction of these approaches.