All talks listed in Pacific Time. Schedule subject to change.

–
Coffee and Check-In
–
Opening Remarks
–
How Do Transformers Work, Part 1
Ankur Moitra (MIT)
–
Break
–
How Do Transformers Work, Part 2
Daniel Hsu (Columbia University)
–
Lunch (on your own)
–
Toward Understanding In-context Learning
Tengyu Ma (Stanford University)
–
Reception
–
Coffee and Check-In
–
LLM Safety, Alignment, and Generalization
Roger Grosse (University of Toronto)
–
Break
–
On large language models and transformers: perspectives from physics, neuroscience, and theory
Surya Ganguli (Stanford University)
–
Lunch
–
Training Large Language Models: Practices and Research Questions
Danqi Chen (Princeton University)
–
Break
–
Understanding and Steering Generative AI Systems
Jacob Steinhardt (UC Berkeley)
–
Coffee and Check-In
–
Challenges in Making LLMs Safe and Robust
Aditi Raghunathan (Carnegie Mellon University)
–
Break
–
AI Safety: Robustness, memorization, and uncertainty quantification
Tatsu Hashimoto (Stanford University)
–
Lunch
–
The Enigma of LLMs: on Creativity, Compositionality, Pluralism, and Paradoxes
Yejin Choi (University of Washington / NVIDIA)
–
Break
–
Discussion Session (Danqi Chen, Yejin Choi, Daniel Hsu, Aditi Raghunathan; moderated by Umesh Vazirani)