Understanding and Improving the Safety of Frontier Models

Abstract: (The talk will be self-contained and no background on LLM Safety/Alignment is required.) This talk provides a foundational overview of recent efforts in industry and academia to improve the safety of frontier models, along with open challenges. It will cover (1) principal approaches to designing red-teaming attacks, (2) in-model and out-of-model methods for enhancing safety, and (3) if time permits, the challenge of catastrophic forgetting in post-training and approaches to continual learning.

Bio: Hamed Hassani is currently a senior research scientist at Google as well as an associate professor of Electrical and Systems Engineering department, the Computer and Information Systems department, and the department of Statistics and Data Science at the University of Pennsylvania. Prior to that, he was a research fellow at Simons Institute for the Theory of Computing (UC Berkeley) affiliated with the program of Foundations of Machine Learning, and a post-doctoral researcher in the Institute of Machine Learning at ETH Zurich. He received a Ph.D. degree in Computer and Communication Sciences from EPFL, Lausanne. He is the recipient of the 2014 IEEE Information Theory Society Thomas M. Cover Dissertation Award, 2015 IEEE International Symposium on Information Theory, 2020 Air Force Office of Scientific Research (AFOSR) Young Investigator Award, 2020 National Science Foundation (NSF) CAREER Award, 2020 Intel Rising Star award, the distinguished lecturer of the IEEE Information Society in 2022-23, the 2023 IEEE Communications Society & Information theory Society Joint Paper Award, and the 2023 ICML-AdvML best paper award. Moreover, he is the recipient of 2023 IEEE Information Theory Society’s James L. Massey Research and Teaching Award for Young Scholars.