Skip to yearly menu bar Skip to main content

In-Person Poster presentation / poster accept

The Onset of Variance-Limited Behavior for Networks in the Lazy and Rich Regimes

Alexander Atanasov · Blake Bordelon · Sabarish Sainathan · Cengiz Pehlevan

MH1-2-3-4 #46

Keywords: [ Scaling Laws ] [ Deep Ensembles ] [ neural tangent kernel ] [ feature learning ] [ Deep Learning and representational learning ]

Abstract: For small training set sizes $P$, the generalization error of wide neural networks is well-approximated by the error of an infinite width neural network (NN), either in the kernel or mean-field/feature-learning regime. However, after a critical sample size $P^*$, we empirically find the finite-width network generalization becomes worse than that of the infinite width network. In this work, we empirically study the transition from infinite-width behavior to this \textit{variance-limited} regime as a function of sample size $P$ and network width $N$. We find that finite-size effects can become relevant for very small dataset sizes on the order of $P^* \sim \sqrt{N}$ for polynomial regression with ReLU networks. We discuss the source of these effects using an argument based on the variance of the NN's final neural tangent kernel (NTK). This transition can be pushed to larger $P$ by enhancing feature learning or by ensemble averaging the networks. We find that the learning curve for regression with the final NTK is an accurate approximation of the NN learning curve. Using this, we provide a toy model which also exhibits $P^* \sim \sqrt{N}$ scaling and has $P$-dependent benefits from feature learning.

Chat is not available.