PaulPauls / llama3_interpretability_sae
A complete end-to-end pipeline for LLM interpretability with sparse autoencoders (SAEs) using Llama 3.2, written in pure PyTorch and fully reproducible.
☆604Updated 2 months ago
Alternatives and similar repositories for llama3_interpretability_sae:
Users that are interested in llama3_interpretability_sae are comparing it to the libraries listed below
- LLM Analytics☆642Updated 4 months ago
- Bayesian Optimization as a Coverage Tool for Evaluating LLMs. Accurate evaluation (benchmarking) that's 10 times faster with just a few l…☆277Updated last week
- A library for making RepE control vectors☆551Updated last month
- Visualize the intermediate output of Mistral 7B☆339Updated 3 weeks ago
- A scientific instrument for investigating latent spaces☆653Updated last week
- Felafax is building AI infra for non-NVIDIA GPUs☆553Updated 3 weeks ago
- Dead Simple LLM Abliteration☆202Updated this week
- An interactive HTML pretty-printer for machine learning research in IPython notebooks.☆389Updated this week
- Textbook on reinforcement learning from human feedback☆438Updated this week
- Things you can do with the token embeddings of an LLM☆1,424Updated 2 weeks ago
- Absolute minimalistic implementation of a GPT-like transformer using only numpy (<650 lines).☆250Updated last year
- Open weights language model from Google DeepMind, based on Griffin.☆620Updated 7 months ago
- Code to train and evaluate Neural Attention Memory Models to obtain universally-applicable memory systems for transformers.☆286Updated 3 months ago
- Official codebase for the paper "Beyond A* Better Planning with Transformers via Search Dynamics Bootstrapping".☆358Updated 8 months ago
- Lightweight Nearest Neighbors with Flexible Backends☆240Updated this week
- A pure NumPy implementation of Mamba.☆219Updated 7 months ago
- A tool to analyze and debug neural networks in pytorch. Use a GUI to traverse the computation graph and view the data from many different…☆280Updated 2 months ago
- A BERT that you can train on a (gaming) laptop.☆210Updated last year
- ☆239Updated 11 months ago
- DiscoGrad - automatically differentiate across conditional branches in C++ programs☆203Updated 5 months ago
- The Fast Vector Similarity Library is designed to provide efficient computation of various similarity measures between vectors.☆379Updated 5 months ago
- An implementation of bucketMul LLM inference☆215Updated 7 months ago
- llama3.np is a pure NumPy implementation for Llama 3 model.☆973Updated 8 months ago
- Minimalistic, extremely fast, and hackable researcher's toolbench for GPT models in 307 lines of code. Reaches <3.8 validation loss on wi…☆342Updated 6 months ago
- Radient turns many data types (not just text) into vectors for similarity search, RAG, regression analysis, and more.☆273Updated last month
- Visualizing the internal board state of a GPT trained on chess PGN strings, and performing interventions on its internal board state and …☆200Updated 3 months ago
- ShellSage saves sysadmins’ sanity by solving shell script snafus super swiftly☆288Updated last week
- A JAX research toolkit for building, editing, and visualizing neural networks.☆1,731Updated 2 months ago