Bibliography (4):

1. Attention Is All You Need

2. MAE: Masked Autoencoders Are Scalable Vision Learners

3. Contrastive Representation Learning: A Framework and Review

4. https://github.com/zinengtang/TVLT