Kiru Lab  /  Mechanistic Interpretability

From Mechanism to Diagnosis

The closing module: turning mechanistic findings into DEM-X evidence, and the standard Kiru holds.

In this module you will

  • Write a mechanistic account that survives adversarial review
  • Assemble an evidence package a stranger could reproduce
  • State the limits of your claim explicitly
Sparse Autoencoders and Dictionary Learning
If superposition hides features in a compact basis, learn a bigger basis in which they separate.
Concept · 40m
From Circuit to Disorder
The closing lesson. Everything in this lab exists to make this document possible.
Bridge · 45m
← Mechanistic Interpretability