Today's Open Source: Tsinghua University and CAS Open-Source LLaVA-UHD (Dec. 9, 2025), PVC Progressive Visual Compression Achieves Efficient Native Resolution Encoding for Multimodal Models

⌘K

F

Today's Open Source: Tsinghua University and CAS Open-Source LLaVA-UHD (Dec. 9, 2025), PVC Progressive Visual Compression Achieves Efficient Native Resolution Encoding for Multimodal Models | BestBlogs.dev