PDE-State Conditioning for Learned Optimization Using Physics-Informed Neural Networkswith Ricardo Vilalta1 versionW_6sz28dc2v1 · current