ArticleNature communications2025

Training a high-performance retinal foundation model with half-the-data and 400 times less compute.

Justin Engelmann et al.PubMed ↗Full text ↗Publisher ↗

No numbers read from the abstract.