Yijia (Zoe) Zheng

I'm a Ph.D. student in Computer Science at Purdue University advised by Prof. Raymond A. Yeh in the Department of Computer Science. I'm also working with David Wipf and Tong He at AWS Shanghai AI Lab. I received my B.Sc. in Data Science at School of Statistics and Management, Shanghai University of Finance and Economics, where I was working with Prof. Yixuan Qiu.

I'm interested in generative AI, explainable AI, and graph machine learning. If you are interested in these topics, feel free to reach out!

Email  /  CV  /  Google Scholar  /  Twitter  /  Github /  Linkedin

profile photo
News

Mar 2023: Invited speaker for Rising Star Lecture Series of A*STAR Center for Frontier AI Research (CFAR) (Link)

Oct 2022: I get NeurIPS 2022 Scholar Award.

Sept 2022: Our paper Manifold Dimensions with Conditional Variational Autoencoders (paper, code) is accepted by NeurIPS 2022!

Aug 2022: I'm on boarding at Purdue as a PhD student. Boiler up!!

Jun 2022: I'm awarded Shanghai Outstanding Graduate and Honours Degree at Shanghai University of Finance and Economics!

Invited Talks

Centre for Frontier AI Research, Singapore. (link, slides)

Research
IMMA: Immunizing text-to-image Models against Malicious Adaptation
Yijia Zheng and Raymond A. Yeh
Project Page, arXiv

We propose IMMA to immunize the pre-trained model by learning model parameters that are difficult for the adaptation methods to fine-tune malicious content.



Learning Manifold Dimensions with Conditional Variational Autoencoders
Yijia Zheng, Tong He, Yixuan Qiu, David Wipf
NeurIPS, 2022. arXiv

We prove that VAE global minima are indeed capable of recovering the correct manifold dimension. Then extend this result to more general CVAEs, demonstrating practical scenarios whereby the conditioning variables allow the model to adaptively learn manifolds of varying dimension across samples

Services

Conference reviewer: NeurIPS, ICLR, CVPR

Awards

NeurIPS 2022, Scholar Award, 2022

Shanghai Outstanding Graduate, 2022

First-class People's Scholarship, 2020, 2021, 2022


Updated in Nov. 2023. Cloned from this popular template.