Dongbin Zhao

WALA achieves a remarkable 75.2% average success rate on RoboCasa, showcasing the power of combining action-labeled and action-free data for robust robot learning.

Tactile dynamics are crucial for contact-rich manipulation, and VT-WAM outperforms existing models by 26.67% to 35.84% by effectively integrating visual and tactile cues.

UCOB achieves unprecedented performance in agentic reinforcement learning by dynamically refining skill usage through credit-aware self-distillation.

Real-time planning for autonomous driving can now achieve superior safety in complex environments by leveraging fast-sampling consistency models for multimodal trajectory generation.

Forget robot-specific fine-tuning: a unified diffusion model can now learn policies across diverse robot embodiments, boosting performance by 15% and opening doors to truly generalizable robotic agents.

Self-play can be dramatically improved by exploiting the "question construction path" it generates as privileged information for self-distillation, leading to 2-3x faster learning.

Zero-shot RL agents can now learn better representations by focusing on dynamics-relevant image regions, leading to state-of-the-art generalization performance.

Fine-tuning generative policies for robotics doesn't have to be a nightmare: POCO offers a stable and efficient RL framework that actually works on real robots.

Agentic RL agents can learn faster and perform better by dynamically maintaining a skill bank that combines high-level task guidance with low-level step-by-step decision support.

Autonomous driving models can learn to avoid accidents *before* they happen by training on expert interventions and anticipating errors.

Ditch expensive, rendering-based RL for autonomous driving: PerlAD uses offline data to train agents in a fast, vector-space pseudo-simulation, outperforming prior methods by 10% on driving score.

Achieve up to 28% better success rates in whole-body mobile manipulation by decoupling base and arm control while intelligently allocating perceptual attention.

Key contribution not extracted.

Imagine training robots to manipulate objects in the real world, but entirely within a high-fidelity, diffusion-based dream.