safety-research / safety-tooling
Inference API for many LLMs and other useful tools for empirical research
☆24Updated last week
Alternatives and similar repositories for safety-tooling:
Users that are interested in safety-tooling are comparing it to the libraries listed below
- ☆12Updated this week
- METR Task Standard☆146Updated last month
- ☆10Updated 8 months ago
- Mechanistic Interpretability Visualizations using React☆235Updated 3 months ago
- Resources for skilling up in AI alignment research engineering. Covers basics of deep learning, mechanistic interpretability, and RL.☆206Updated last year
- ControlArena is a suite of realistic settings, mimicking complex deployment environments, for running control evaluations. This is an alp…☆28Updated this week
- Improving Steering Vectors by Targeting Sparse Autoencoder Features☆17Updated 4 months ago
- ☆213Updated 5 months ago
- Open source replication of Anthropic's Crosscoders for Model Diffing☆48Updated 5 months ago
- (Model-written) LLM evals library☆18Updated 3 months ago
- Machine Learning for Alignment Bootcamp☆72Updated 2 years ago
- ☆17Updated last month
- Vivaria is METR's tool for running evaluations and conducting agent elicitation research.☆85Updated this week
- The nnsight package enables interpreting and manipulating the internals of deep learned models.☆526Updated this week
- Delphi was the home of a temple to Phoebus Apollo, which famously had the inscription, 'Know Thyself.' This library lets language models …☆165Updated this week
- ☆66Updated 4 months ago
- ☆43Updated 2 months ago
- A collection of different ways to implement accessing and modifying internal model activations for LLMs☆14Updated 5 months ago
- Using sparse coding to find distributed representations used by neural networks.☆225Updated last year
- Sparse Autoencoder for Mechanistic Interpretability☆233Updated 8 months ago
- ☆264Updated last month
- ☆78Updated last week
- Code for my NeurIPS 2024 ATTRIB paper titled "Attribution Patching Outperforms Automated Circuit Discovery"☆30Updated 9 months ago
- Tools for studying developmental interpretability in neural networks.☆86Updated 2 months ago
- ☆32Updated 4 months ago
- ☆30Updated 11 months ago
- ☆114Updated 7 months ago
- A library for efficient patching and automatic circuit discovery.☆59Updated last month
- LLM experiments done during SERI MATS - focusing on activation steering / interpreting activation spaces☆91Updated last year
- ☆151Updated this week