Abstract

The Cambridge ELLIS Unit Seminar Series holds talks by leading researchers in the area of machine learning and AI. Our next speaker of 2023 will be Dr. Jasper Snoek. Details of his talk can be found below.

Title: “Kernel (GP) Regression with Infinite-Width Neural Networks on Millions of Examples “

Abstract: I’ll talk about some of our recent work on scaling up Gaussian process regression with expressive kernels on structured data.  Neural kernels have drastically increased performance on diverse and nonstandard data modalities but require significantly more compute, which previously limited their application to smaller datasets. In this work, we address this by massively parallelizing their computation across many GPUs. We combine this with a distributed, preconditioned conjugate gradients algorithm to enable kernel regression at a large scale (i.e. up to five million examples). Using this approach, we study scaling laws of several neural kernels across many orders of magnitude for the CIFAR -5m dataset. Using data augmentation to expand the original CIFAR -10 training dataset by a factor of 20, we obtain a test accuracy of 91.2\% (SotA for a pure kernel method). Moreover, we explore neural kernels on other data modalities, obtaining results on protein and small molecule prediction tasks that are competitive with SotA methods.