Biography
I am a research scientist at Google Deepmind. Prior to that, I received my Ph. D. from UTCS advised by Raymond J. Mooney. I received my BEng. degree from the Department of Automation supervised by Prof. Xiangyang Ji at Tsinghua University in 2017.
My most recent research interests are in enhancing the capabilities of image generation models on info-seeking (world knowledge) queries. I work on (1) utilizing search signals during the pre/post-training phases as well as during inference for image generation, and (2) enhancing the factual accuracy of images produced in response to info-seeking queries.
Interests
- Info-seeking Image Generation
- Scaling Laws for Image Generation
Education
PhD in Artificial Intelligence, 2017 - 2022
UT Austin
BEng in Automation, 2013 - 2017
Tsinghua University
Selected Publications and Projects
Please see my google scholar for full publications and preprints.
(2025).
Gemini 2.5 Flash Image.
AI Studio.
(2025).
Gemini 2.5:Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities.
Arxiv.
(2024).
Distilling vision-language models on millions of videos.
CVPR 2024.
(2024).
Omni-SMoLA:Boosting Generalist Multimodal Models with Soft Mixture of Low-rank Experts.
CVPR 2024 (Highlight Poster).
(2023).
CausalLM Is Not Optimal for In-Context Learning.
ICLR 2024.