Sparse Autoencoders Python Packages

glassbox-mech-interp

Open-source EU AI Act Annex IV compliance toolkit. Mechanistic interpretability + circuit discovery for transformers. One function call generates a court-ready evidence package

reptimeline

Track how discrete representations evolve during neural network training — lifecycle events, phase transitions, ontology discovery, and causal verification

tiny-dashboard

A tiny easily hackable implementation of a feature dashboard.

hypothesaes

Hypothesizing interpretable relationships in text datasets using sparse autoencoders.

glassbox-mcp

Open-source EU AI Act Annex IV compliance toolkit. Mechanistic interpretability + circuit discovery for transformers. One function call generates a court-ready evidence package