Introduction to Interpretable AI: Superposition and Sparse Autoencoders

During the event Khalil will present us in introduction to interpretable AI and sparse auto-encoder.

SUMMARY

This talk introduces the field of Interpretable AI by exploring how neural networks store concepts in "superposition" and how Sparse Autoencoders can be used to extract and understand these hidden features.

The presentation will be held at the grande bibliotheque

475 Boul. de Maisonneuve E, Montréal, QC H2L 5C4

room 2.130.1

The goal of this "informal" event is to learn about AI and discuss it together with the host.