Jingren Zhou

WanSong achieves high-fidelity song generation in a single diffusion pass, revolutionizing how we think about music creation models.

Video can be reimagined as a dynamic interplay of stable contexts and evolving events, revolutionizing real-time interaction capabilities in AI.

Higher resolution in real-time audio-visual interactions can be achieved without sacrificing latency, enabling clearer agent representation in conversations.

EAPO revolutionizes LLM reasoning by dynamically integrating prior experiences, leading to consistent performance gains over traditional RLVR methods.

ForeAgent outperforms existing deepfake detection methods by 16.41% while continuously evolving its reasoning capabilities through self-reflection and high-quality sample generation.

EDA not only corrects the current memory write but also actively removes outdated information, leading to superior performance in long-context scenarios.

Sub-second duplex audio-visual communication is now achievable with a single, unified model that eliminates the latency of traditional cascaded systems.

Qwen-AgentWorld achieves unprecedented simulation fidelity, outperforming existing models and enabling scalable agentic reinforcement learning across diverse real-world environments.

Bypassing final-layer perturbations can significantly enhance reasoning capabilities in aligned LLMs, achieving better performance with zero memory overhead.

LLMs can achieve remarkable out-of-distribution generalization by learning to self-update their context through a novel reinforcement learning framework.

Qwen-RobotNav redefines navigation by allowing real-time reconfiguration of strategies, achieving unprecedented flexibility and performance across diverse tasks.

Qwen-RobotManip achieves a 20% relative improvement over the previous state-of-the-art in robotic manipulation, showcasing unprecedented generalization capabilities from diverse, open-source datasets.

Language-driven video generation in Qwen-RobotWorld achieves unprecedented accuracy in predicting robotic actions, outperforming existing models across key benchmarks.

MTP acceptance rates can be dramatically improved by addressing entropy fluctuations, leading to up to 1.8x faster RL training.

Rethinking few-step distillation reveals that the training pipeline's organization is as crucial as the distillation objectives themselves.