4 accepted papers
Adam outperforms SGD when training language models. Yet this advantage is not well-understood theoretically -- previous convergence analysis for Adam and SGD mainly focuses on the number of steps $T$ and is already minimax-optimal in non-convex cases, which are both $\widetilde{O}(T^{-1/4})$. In th…
We present a theoretical explanation of the “grokking” phenomenon (Power et al., 2022), where a model generalizes long after overfitting, for the originally-studied problem of modular addition. First, we show that early in gradient descent, so that the “kernel regime” approximately holds, no permuta…
Empirical neural tangent kernels (eNTKs) can provide a good understanding of a given network's representation: they are often far less expensive to compute and applicable more broadly than infinite-width NTKs. For networks with $O$ output units (e.g. an $O$-class classifier), however, the eNTK on $N…
We propose a new method for approximating active learning acquisition strategies that are based on retraining with hypothetically-labeled candidate data points. Although this is usually infeasible with deep networks, we use the neural tangent kernel to approximate the result of retraining, and prove…