Tensorflow: intercepting, changing gradient and continuing backpropagation

Is it possible to intercept the inverse gradient from some arbitrary layer, change its values ​​and continue unfolding back to the beginning of the network, updating the inverse gradients of all previous layers based on the changed gradient values ​​that you specified?

I know that you can directly modify the gradients themselves before applying the update , but as far as I know it would only change the gradients of the specified layer without propagating to the gradients of the previous layers.