A simple framework for contrastive learning of visual representations (SimCLR).¶
Chen, T., Kornblith, Norouzi, & Hinton, G. (2020). A simple framework for contrastive learning of visual representations (SimCLR). Proceedings of the 37th International Conference on Machine Learning (ICML), 1597-1607.
Cited by¶
1 citation across 1 artifact.
Each citation links to the sentence it supports in the citing article.
Primes¶
- Contrast
- Chen, Kornblith, Norouzi, and Hinton (2020), in their SimCLR framework, made this pattern explicit for representation learning: a model is trained to pull positive pairs together and push negative pairs apart in embedding space, using contrast itself as the supervisory signal.
This sourceTrains models to maximize agreement between augmented views of the same image (positive pair) and minimize agreement with views of other images (negative pairs) — supports the claim that contrast itself serves as the supervisory signal for representation learning. (Preprint: arXiv:2002.05709.)
Supported in partVerified against the publisher's abstract
The SimCLR abstract supports a contrastive learning framework for visual representations, but does not spell out pulling positives together and pushing negatives apart.
“This paper presents SimCLR: a simple framework for contrastive learning of visual representations.”
- Chen, Kornblith, Norouzi, and Hinton (2020), in their SimCLR framework, made this pattern explicit for representation learning: a model is trained to pull positive pairs together and push negative pairs apart in embedding space, using contrast itself as the supervisory signal.
Verification¶
Does it exist? Not checked yet. This entry carries no identifier to resolve. It was extracted from the citation as written in the article, normalized, and deduplicated against the rest of the registry.
Does it back the claim? Read against the text for 1 of 1 citation: 1 supported in part. Each verdict is shown under its citation below, with what in the work backs the sentence.
Was it audited? Yes. A second, independent pass read the citation against the article text and recorded a verdict.
Support is checked per citation rather than per work — the same source can be cited soundly in one article and wrongly in another. Per-citation recording began recently, so a citation with no recorded check is a gap in the record rather than evidence it went unchecked.
See how references were verified.
Registry ID ref:e9f0fb7ec1ac · see in the full table