PyPI page
Home page
Author:
Ajay Pravin Mahale
License:
MIT AND LicenseRef-BUSL-1.1
Summary:
Mechanistic interpretability for transformers: causal circuit discovery, measured faithfulness (sufficiency, comprehensiveness, F1) and draft EU AI Act Annex IV documentation. Includes an AI-system investigation layer. Dual-licensed (MIT core + BSL 1.1 compliance engine and investigation layer).
Latest version:
4.6.1
Required dependencies:
einops
|
numpy
|
scipy
|
torch
|
transformer-lens
Optional dependencies:
fastapi
|
huggingface-hub
|
ipython
|
ipywidgets
|
mlflow
|
opentelemetry-exporter-otlp
|
opentelemetry-sdk
|
pydantic
|
pytest
|
reportlab
|
sae-lens
|
uvicorn
Downloads last day:
10
Downloads last week:
139
Downloads last month:
105
7-day smoothing