Stars
Give your houseplants a voice with ChatGPT.
Personalized Steering of Large Language Models: Versatile Steering Vectors Through Bi-directional Preference Optimization
LLM Transparency Tool (LLM-TT), an open-source interactive toolkit for analyzing internal workings of Transformer-based language models. *Check out demo at* https://huggingface.co/spaces/facebook/l…
Unbearably fast near-real-time pure-Python runtime-static type-checker.
Type annotations and runtime checking for shape and dtype of JAX/NumPy/PyTorch/etc. arrays. https://docs.kidger.site/jaxtyping/
Processed / Cleaned Data for Paper Copilot
A complete end-to-end pipeline for LLM interpretability with sparse autoencoders (SAEs) using Llama 3.2, written in pure PyTorch and fully reproducible.
Tools for understanding how transformer predictions are built layer-by-layer
A TinyStories LM with SAEs and transcoders
Sparsify transformers with SAEs and transcoders
Delphi was the home of a temple to Phoebus Apollo, which famously had the inscription, 'Know Thyself.' This library lets language models know themselves through automated interpretability.
Provenance and caching library for python functions, built for creating lightweight machine learning pipelines
Funes the Memorious decorates functions to memoize their computations permanently to disk.
Training Sparse Autoencoders on Language Models
A reactive notebook for Python — run reproducible experiments, query with SQL, execute as a script, deploy as an app, and version with git. Stored as pure Python. All in a modern, AI-native editor.
Code for "Is This the Subspace You Are Looking for? An Interpretability Illusion for Subspace Activation Patching"
Erasing concepts from neural representations with provable guarantees
A library for mechanistic interpretability of GPT-style language models
Sparse probing paper full code.
Playing around with transformers, on the path towards CCA analysis
Platform for testing and improving ML continuously.