Topic
Qwen3 Vl
2 episodes
-
VideoChat3:Fully Open Video MLLM for Efficient and Generalist Video Understanding
VideoChat3 is a fully open-source video multimodal LLM (4B parameters) that tackles three concrete problems: generalization across short/long/streaming video, computational efficiency for video token explosion, and reproducibility through complete open-sourcing. The core innovation is I3D-ViT (Inflated 3D Vision Transformer) plus adaptive frame resolution, which compresses spatiotemporal redundancy early in the pipeline instead of treating each frame as an independent image. Three curated datasets (2M academic + 116K long-form + 617K streaming = 3M samples total) and multi-stage curriculum learning enable the model to handle diverse video scenarios. Jessica sees a shippable foundation for real-world video apps; Cathy pushes on whether the efficiency gains hold under production load and whether the data pipeline's scale claim is reproducible.
-
PixelRAG beats text parsers, cuts agent costs 10x
Justy and Cody dissect PixelRAG, a new research system that skips text parsing entirely by feeding rendered webpage screenshots directly to vision-language models. They break down the three specific failure modes of traditional parsers (parser loss, rank loss, reader loss) and discuss whether the 10x cost reduction and accuracy gains hold up against the engineering reality of managing image indices.