How to continue BP using intermediate gradients

If you have a reference to the activation Tensor somewhere, you can access its .grad attribute. If not, there’s a little more work you would need to do, but still possible.