Kang Chen

J64 reveals hidden reasoning states that can significantly boost model accuracy and decision-making, while R64 provides a lightweight, effective proxy for deployment.

Current attention mechanisms misjudge the relevance of visual information, risking the loss of crucial context in multimodal dialog systems.

STEAM redefines how robots learn from mixed-quality data, achieving up to 59% higher success rates in real-world tasks by effectively identifying reliable progress.

Identical token IDs in sparse MoE models can lead to different expert outputs, revealing a nuanced structure that enhances reasoning control through innovative routing techniques.

LaWAM achieves up to 24x lower latency than traditional pixel-space models while maintaining state-of-the-art performance in robot control tasks.

Forget static datasets – RL-based co-training unlocks +20% real-world VLA performance by interactively leveraging simulation while preserving real-world capabilities.