CyLab Seminar: Lei Li
January 22, 2024
12:00 p.m. ET
Zoom or Panther Hollow 4105 (4th Floor of the CIC)
January 22, 2024
12:00 p.m. ET
Zoom or Panther Hollow 4105 (4th Floor of the CIC)
*Please note: this CyLab seminar is open only to partners and Carnegie Mellon University faculty, students, and staff.
Speaker:
Lei Li
Assistant Professor
Carnegie Mellon University Language Technologies Institute
Talk Title:
Is It Generated by AI?
Attacks and Robust Watermarking for Generative AI
Abstract: Generative AI have demonstrated incredible power, but the synthetic generation also challenges authenticity of the content. In this talk, we will address two fundamental questions regarding responsible use of generative AI: (1) How can we accurately identify AI-generated text and images? (2) What measures can safeguard the intellectual property of GenAI? We will first present a strong image watermark attacking method and a paraphrasing-based text watermark attacking approach that are capable of removing watermarks. We will introduce two robust watermarking techniques designed for text and models. Our discussion will encompass the theoretical underpinnings that ensure the correctness of watermark detection, along with robustness against evasion attacks. These findings establish a solid technical groundwork for policymakers, legal professionals, and generative AI practitioners alike. (joint work with Xuandong Zhao, Yu-Xiang Wang, Prabhanjan Ananth, Kexun Zhang)
Bio: Lei Li is an Assistant Professor in the Language Technologies Institute at Carnegie Mellon University. His research interests are machine translation, trustworthy LLMs, and AI drug discovery. He received his Ph.D. from the Carnegie Mellon University School of Computer Science in 2011. He is a recipient of ACL 2021 Best Paper Award, CCF Young Elite Award in 2019, CCF distinguished speaker in 2017, Wu Wen-tsün AI prize in 2017, and 2012 ACM SIGKDD dissertation award (runner-up), and is recognized as Notable Area Chair of ICLR 2023. Previously, he was an associate professor (tenured) at UC Santa Barbara. Prior to that, he was the Founding Director of ByteDance AI Lab. He led and developed ByteDance’s machine translation system VolcTrans and AI writing system Xiaomingbot, and many of his algorithms have been deployed in products (Toutiao, Douyin, TikTok, Lark), serving over a billion users.
November 8 2024
8:30 AM - 5:00 PM ET
CyLab Security and Privacy Institute
Industry-Academia Partnership (IAP) Workshop — AI and Security in the Cloud
Gates and Hillman Centers, room 6115
November 11 2024
12:00 PM ET
CyLab Security and Privacy Institute
Zoom or CIC room 4105, Panther Hollow
November 18 2024
12:00 PM ET
CyLab Security and Privacy Institute
Zoom or CIC room 4105, Panther Hollow