Xuanyu Zheng

Pre-indexing hours of video is a computational dead end—growing an online, query-driven textual memory tree during inference matches offline accuracy without the upfront processing tax.

VideoRover unifies video reasoning and external knowledge retrieval, achieving competitive performance with fewer resources than larger models.

Achieving lossless processing of 256K contexts, Keye-VL-2.0 transforms how we approach long-video understanding and agentic intelligence.

Key contribution not extracted.

Ditching modular architectures unlocks surprisingly competitive vision-language performance, proving that end-to-end pixel-to-word models can rival traditional approaches at scale.