Seminar Series: Security

Abdelnabi will discuss his work on evaluating multi-agent negotiations and how that can be beneficial to test reasoning and create evolving, dynamic benchmarks.

This presentation demonstrates how red teaming uncovers critical vulnerabilities in AI agents that challenge assumptions about safe deployment.

How can researchers use sensitive datasets for machine learning and statistics without compromising the privacy of the individuals who contribute their data?

In this talk, Rando will introduce the notion of a tokenizer shortcut that approximates tokenization with a continuous function and enables continuous...

In this talk, Anshuman Suri will challenge these assumptions and demonstrate that optimal membership inference does require parameter access.

Norman Mu is a Member of Technical Staff at xAI, focusing on AI safety. He earned his Ph.D. in Computer Science from the University of California, Berkeley...