Physics Department Colloquium

Scaling towards AGI

When
-

Abstract: In this talk, I will take you on a tour of large language models, tracing their evolution from Recurrent Neural Networks (RNNs) to the Transformer architecture. We will explore how Transformers elegantly sidestep the vanishing and exploding gradient issues that plagued RNNs. I will introduce neural scaling laws—empirical relationships reminiscent of scaling behaviors common in physics—that predict how model performance improves with increased computational investment.

From Mars Sample Return to Enceladus plume missions: finding habitable environments and life across the solar system

When
-

Abstract:  Planetary exploration has unveiled environments that could support life today, or in the past, or contain the ingredients of life. Mars was once habitable; Enceladus’ ocean is today, Europa is a question mark, and Bennu contains most of the basic key monomers of life in abiotic form. I will discuss how this all fits together.