Junyuan Hong
@hjy836
Postdoc @VITAGroupUT @UTAustin @MLFoundations. Ph.D @michiganstateu . Ex @SonyAI_global intern. Interests: trustworthy ML, federated learning, privacy, LLM
08-09-2013 02:58:04
237 Tweets
702 Followers
2,0K Following
![[Finetuning can amplify the privacy risks of Generative AI (Diffusion Models)] Last week, I was honored to give a talk at the Good Systems Symposium (gssymposium2024.splashthat.com), where I shared our recent work on the π¨ privacy risks of Generative AI via finetuning. Our leading](https://pbs.twimg.com/media/GKsUdhmXQAAYSMJ.png "Junyuan Hong (@hjy836) on Twitter photo 2024-04-09 03:25:23 [Finetuning can amplify the privacy risks of Generative AI (Diffusion Models)] Last week, I was honored to give a talk at the Good Systems Symposium (gssymposium2024.splashthat.com), where I shared our recent work on the π¨ privacy risks of Generative AI via finetuning. Our leading")
Work is led by Zhangheng Li. Amazing collaboration with Bo Li Secure Learning Lab (SLL) (UChicago) and Zhangyang Wang VITA Group (UT Austin).