Modelpedia
Work in progress
About
Findings
Models
Concepts
Methods
Datasets
Sources
Light
Dark
Is Large-scale Pretraining the Secret to Good Domain Generalization?
2025-01-22
· ICLR 2025 Poster ·
anchor
Findings
IC-553
OpenCLIP ViT-B/16's image-text alignment score is a strong predictor of domain generalization accuracy, while perceptual similarity to LAION-400M pre-training data is a weaker predictor