REVIEW 2 cited by
Contrastive Learning Can Find An Optimal Basis For Approximately View-Invariant Functions
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Contrastive learning is a powerful framework for learning self-supervised representations that generalize well to downstream supervised tasks. We show that multiple existing contrastive learning methods can be reinterpreted as learning kernel functions that approximate a fixed positive-pair kernel. We then prove that a simple representation obtained by combining this kernel with PCA provably minimizes the worst-case approximation error of linear predictors, under a straightforward assumption that positive pairs have similar labels. Our analysis is based on a decomposition of the target function in terms of the eigenfunctions of a positive-pair Markov chain, and a surprising equivalence between these eigenfunctions and the output of Kernel PCA. We give generalization bounds for downstream linear prediction using our Kernel PCA representation, and show empirically on a set of synthetic tasks that applying Kernel PCA to contrastive learning models can indeed approximately recover the Markov chain eigenfunctions, although the accuracy depends on the kernel parameterization as well as on the augmentation strength.
Forward citations
Cited by 2 Pith papers
-
Visual Pre-Training on Unlabeled Images using Reinforcement Learning
Casting image-crop consistency as temporal-difference value learning improves visual representations on unlabeled web, scene, and video data.
-
Representation Learning for Equivariant Inference with Guarantees
An equivariant representation learning framework estimates conditional distributions through a block-diagonal conditional expectation operator, with non-asymptotic sample-complexity guarantees that improve with the sy...
Discussion (0). Sign in to comment.