mrfusion on Every Model Learned by Gradient Descent Is Approximately a Kernel Machine · Hackerly

Comment on Every Model Learned by Gradient Descent Is Approximately a Kernel Machine

Does this include transformers?