Pith. sign in

REVIEW 1 cited by

Contrastive learning, multi-view redundancy, and linear models

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

arxiv 2008.10150 v2 pith:RTP7CQV6 submitted 2020-08-24 cs.LG stat.ML

Contrastive learning, multi-view redundancy, and linear models

classification cs.LG stat.ML
keywords learningcontrastiveviewsapproachdatalinearmulti-viewself-supervised
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

Self-supervised learning is an empirically successful approach to unsupervised learning based on creating artificial supervised learning problems. A popular self-supervised approach to representation learning is contrastive learning, which leverages naturally occurring pairs of similar and dissimilar data points, or multiple views of the same data. This work provides a theoretical analysis of contrastive learning in the multi-view setting, where two views of each datum are available. The main result is that linear functions of the learned representations are nearly optimal on downstream prediction tasks whenever the two views provide redundant information about the label.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. MABLE: Masked Autoencoding with Bi-Lipschitz Decoding for Embeddings and Graph Metric Learning

    cs.LG 2026-07 conditional novelty 5.5

    MABLE learns stable node and graph embeddings on heterogeneous geospatial graphs via masked autoencoding, bi-Lipschitz decoding, and fixed cosine alignment/uniformity losses without discriminators or hard negatives.