wesg52

mamba

Mamba SSM architecture

kedro

Kedro is a toolbox for production-ready data science. It uses software engineering best practices to help you create data engineering and data science pipelines that are reproducible, maintainable, and modular.

transformer-debugger

leafmap

A Python package for interactive mapping and geospatial analysis with minimal coding in a Jupyter environment

anthropic-sdk-python

awesome-neural-geometry

A curated collection of resources and research related to the geometry of representations in the brain, deep networks, and beyond

representation-engineering

Representation Engineering: A Top-Down Approach to AI Transparency

pyvene

Stanford NLP Python Library for Understanding and Improving PyTorch Models via Interventions

redun

Yet another redundant workflow engine

torchlens

Package for extracting and mapping the results of every single tensor operation in a PyTorch model in one line of code.

cola

Compositional Linear Algebra

SAELens

Training Sparse Autoencoders on Language Models

nnsight

The nnsight package enables interpreting and manipulating the internals of deep learned models.

inseq

Interpretability for sequence generation models 🐛 🔍

Awesome-Interpretability-in-Large-Language-Models

This repository collects all relevant resources about interpretability in LLMs

world-models

Extracting spatial and temporal world models from LLMs

Automatic-Circuit-Discovery

sparse_autoencoder

Sparse Autoencoder for Mechanistic Interpretability

sae_vis

Create feature-centric and prompt-centric visualizations for sparse autoencoders (like those from Anthropic's published research).

dictionary_learning

Awesome-LLM-Interpretability

A curated list of LLM Interpretability related material - Tutorial, Library, Survey, Paper, Blog, etc..

sleeper-agents-paper

Contains random samples referenced in the paper "Sleeper Agents: Training Robustly Deceptive LLMs that Persist Through Safety Training".

devinterp

Tools for studying developmental interpretability in neural networks.

modeldiff

ModelDiff: A Framework for Comparing Learning Algorithms

sparse-probing-paper

Sparse probing paper full code.

universal-neurons

Universal Neurons in GPT2 Language Models

elk-generalization

Investigating the generalization behavior of LM probes trained to predict truth labels: (1) from one annotator to another, and (2) from easy questions to hard

edge-attribution-patching

Code for my NeurIPS 2024 ATTRIB paper titled "Attribution Patching Outperforms Automated Circuit Discovery"

ActivationDirectionAnalysis

llmICL