Jingren Zhou

Qwen-VLA introduces a unified embodied foundation model that extends vision-language understanding to continuous action generation, enabling robust, multi-task generalization across diverse robotic tasks and embodiments.

This paper introduces WanSong, a diffusion-based model for long-form, commercial-grade song generation that directly generates high-fidelity, multilingual songs up to 5 minutes and outputs dual stems, enabling faster inference and customization.

Libra is a load balancing approach for long-context LLM training that groups packed sequences into fixed-size pools and reduces attention workload variance, improving end-to-end throughput and straggler-attention speedup.