Fetching the paper…

TCT: Convexifying Federated Learning using Bootstrapped Neural Tangent Kernels · Around