[eismcc]

Has anyone tried using another NN as the “gradient” function? With the hope of making it less “blind?”

[rolisz]

DeepMind has a thing called synthetic gradients: https://www.deepmind.com/blog/decoupled-neural-interfaces-us....