Andy Arditi's repositories
Language:SCSS000
Language:HTML000
CircuitsVis
Mechanistic Interpretability Visualizations using React
Language:Jupyter NotebookMIT000
Language:PythonMIT000
Language:Jupyter NotebookMIT000
mi
Repo to track miscellaneous mi stuff
Language:Jupyter Notebook000
path_patching
Implementation of path patching & activation patching (will eventually add to TransformerLens).
Language:Python000
SycophancySteering
Modulating sycophancy in llama-2 via activation steering
Language:Python000
TransformerLens
A library for mechanistic interpretability of GPT-style language models
Language:PythonMIT000