AWhat changed
Replaced globally normalized softmax image-text contrastive loss with pairwise sigmoid classification, decoupling the objective from global batch normalization and enabling efficient distributed scaling.
Scientific scope
Replaced globally normalized softmax image-text contrastive loss with pairwise sigmoid classification, decoupling the objective from global batch normalization and enabling efficient distributed scaling. Paired image-text pretraining and sigmoid classification are inherited. No-global-normalizer does not remove negative-pair computation, communication or all batch effects. Original v1 excludes later recipe gains.
Evidence
The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Paired image-text pretraining and sigmoid classification are inherited. No-global-normalizer does not remove negative-pair computation, communication or all batch effects. Original v1 excludes later recipe gains.
Contributors
- Source-listed authors and technical contributors · SCIENCE: Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.
Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.
Full scientific record · Browse contribution grades · Request a correction