← Curriculum
Module 01 Week 1–2

Foundations: Activations, Probes & Features

The groundwork — residual streams, linear probes, and the linear representation hypothesis. Everything later modules build on.

foundationsprobes

Placeholder module — replace with the real syllabus

Learning goals

  • Read and manipulate residual-stream activations.
  • Train and interpret a linear probe.
  • State the linear representation hypothesis and its limits.

Readings

  • Placeholder reading 1 — swap in the real link.
  • Placeholder reading 2 — swap in the real link.

Exercises

  1. Extract activations from an open-weights model and probe for a known concept.
  2. Report where the probe succeeds and where it fails.