Skip to the content.

This is the webpage for the 2023 conference “Singularities against the Singularity” in Berkeley, CA. For further information see the announcement and the page for Registration.

The Primer

The aim of the Primer is to give a general introduction to Singular Learning Theory (SLT) and related areas of mathematics and physics, with the aim of providing a foundation for theoretical and experimental work on AI alignment. More concretely, we aim to explain the Free Energy Formula derived by Watanabe, what its terms mean, how to apply it to understand the phase structure of a learning machine, and how to derive intuition for the resulting picture from physics.

Time Monday Tuesday Wednesday Thursday Friday
9:00-10:00 Welcome / SLT High 1 SLT High 2 SLT High 3 SLT High 4 SLT High 5
10:30-11:00 break break break break break
11:00-12:00 SLT Low 1 SLT Low 2 SLT Low 3 SLT Low 4 SLT Low 5
12:00-1:30 lunch lunch lunch lunch lunch
1:30-3:00 Physics 1 Physics 2 Physics 3 Physics 4 Physics 5
3:00-3:30 break break break break break
3:30-4:30 Alignment 1 Alignment 2 Mech interp 1 Mech interp 2 Wrapup

Each day is organised around a general theme, with the final day culminating in a sketch of the derivation of the Free Energy Formula.

SLT High Road

The SLT “high road” explains the conceptual toolkit and how to use it to reason about learning machines, leaving the proofs and details for later (“just tell me why it’s useful to know this”).

SLT Low Road

The SLT “low road” looks at detailed examples and calculations and sketches of how the mathematical theory fits together (“show me how it works in an example”).


Alignment and Mechanistic Interpretability


Week 2

The second week of the workshop is for discussing open problems, collaboration and more mathematical details beyond the introductions in the first week.