AI RESEARCH
Scaling Pre-training to One Hundred Billion Data for Vision Language Models
arXiv CS.CV
•
ArXi:2502.07617v2 Announce Type: replace We provide an empirical investigation of the potential of pre-