CORNN: Convex optimization of recurrent neural networks for rapid inference of neural dynamics Fatih Dinc Department of Applied Physics Stanford University Stanford, CA94305 Adam Shai