Scholar

Zhendong Wang

Google Scholar ID: lRiIjhcAAAAJ

University of Texas at Austin

Reinforcement LearningGenerative Models

Citations & Impact

All-time

Citations

1,715

H-index

i10-index

Publications

Co-authors

list available

Contact

Publications

20 items

Browse publications on Google Scholar (top-right) ↗

Resume (English only)

Academic Achievements

- Publications:
* One-Step Diffusion Policy: Fast Visuomotor Policies via Diffusion Distillation, published on Nvidia Website
* Diffusion Policies creating a Trust Region for Offline Reinforcement Learning, NeurIPS 2024
* Relative Preference Optimization: Enhancing LLM Alignment through Contrasting Responses across Identical and Diverse Prompts, ArXiv
* In-Context Learning Unlocked for Diffusion Models, NeurIPS 2023
* Patch Diffusion: Faster and More Data-Efficient Training of Diffusion Models, NeurIPS 2023
* Diffusion Policies as an Expressive Policy Class for Offline Reinforcement Learning, ICLR 2023
* Diffusion-GAN: Training GANs with Diffusion, ICLR 2023
* Probabilistic Conformal Prediction Using Conditional Random Samples, ICLR 2023
- Open-sourced code for multiple projects on GitHub

Research Experience

Education

- Ph.D. in Statistics and Data Science, University of Texas at Austin, supervised by Prof. Mingyuan Zhou
- Master’s degree in Data Science, Columbia University
- Bachelor’s degree, Tongji University (China), spent a year as an exchange student at the University of California, Berkeley

Background

- Research Interests: deep generative models, reinforcement learning and their applications
- Professional Fields: diffusion models, GANs, online/offline RL, imitation learning, policy optimization, multimodal large language models, conformal predictions, and robust modeling techniques
- Brief Introduction: Currently a Senior Researcher at Microsoft GenAI Team, open to collaborations, discussions, and exploring new opportunities.

Miscellany