Texonom
Texonom
/
Engineering
Engineering
/Data Engineering/Artificial Intelligence/AI Problem/AI Alignment/Explainable AI/Interpretable AI/Mechanistic interpretability/Activation Engineering/Transformer Lens/
TransformerLens 2.0
Search

TransformerLens 2.0

Creator
Creator
Seonglae ChoSeonglae Cho
Created
Created
2025 Mar 6 22:17
Editor
Editor
Seonglae ChoSeonglae Cho
Edited
Edited
2025 Mar 6 22:17
Refs
Refs
 
 
 
 
 
 
 
TransformerLens 2.0 - TransformerLens Documentation
I am very happy to announce TransformerLens now has a 2.0 release! If you have been using recent versions of TransformerLens, then the good news is that not much has changed at all. The primary motivation behind this jump is to transition the project to strictly following semantic versioning as described here. At the last minute we did also remove the recently added HookedSAE, so if you had been using that, I would direct you to Joseph Bloom’s SAELens. Bundled with this major version change are also a handful of internal modifications that only affect contributors.
TransformerLens 2.0 - TransformerLens Documentation
https://transformerlensorg.github.io/TransformerLens/content/news/release-2.0.html
 
 

Recommendations

Texonom
Texonom
/
Engineering
Engineering
/Data Engineering/Artificial Intelligence/AI Problem/AI Alignment/Explainable AI/Interpretable AI/Mechanistic interpretability/Activation Engineering/Transformer Lens/
TransformerLens 2.0
Copyright Seonglae Cho