Andrew Stephen Fletcher
Blog
Projects
Categories
All
(1)
linear probes
(1)
mechanistic interpretability
(1)
residual stream
(1)
Blog
Documenting my learning of AI Safety.
The Residual Stream
mechanistic interpretability
residual stream
linear probes
This post discusses the residual stream in transformer models and builds towards using it for two examples of mechanistic interpretability: Logit Lens
(nostalgebraist 2020)
a…
Feb 10, 2026
No matching items
