Comment on Every Model Learned by Gradient Descent Is Approximately a Kernel MachineparentComments−smallnamespace5yDefining ‘similar’ robustly is the meat of the problem, and what we’re finding deep NNs to do well.
Comments
Defining ‘similar’ robustly is the meat of the problem, and what we’re finding deep NNs to do well.