TransformerLensOrg / CircuitsVis
Mechanistic Interpretability Visualizations using React
☆198Updated 4 months ago
Related projects ⓘ
Alternatives and complementary repositories for CircuitsVis
- ☆188Updated last month
- Create feature-centric and prompt-centric visualizations for sparse autoencoders (like those from Anthropic's published research).☆157Updated last month
- Using sparse coding to find distributed representations used by neural networks.☆184Updated last year
- Training Sparse Autoencoders on Language Models☆469Updated this week
- The nnsight package enables interpreting and manipulating the internals of deep learned models.☆402Updated this week
- ☆105Updated last month
- Sparse Autoencoder for Mechanistic Interpretability☆188Updated 4 months ago
- ☆108Updated last year
- ☆145Updated 3 weeks ago
- Resources for skilling up in AI alignment research engineering. Covers basics of deep learning, mechanistic interpretability, and RL.☆200Updated 9 months ago
- ☆107Updated this week
- Tools for studying developmental interpretability in neural networks.☆75Updated last week
- Steering vectors for transformer language models in Pytorch / Huggingface☆65Updated last month
- ☆98Updated 3 months ago
- Steering Llama 2 with Contrastive Activation Addition☆97Updated 5 months ago
- Sparse autoencoders☆342Updated last week
- Mechanistic Interpretability for Transformer Models☆49Updated 2 years ago
- ☆44Updated this week
- Tools for understanding how transformer predictions are built layer-by-layer☆430Updated 5 months ago
- Keeping language models honest by directly eliciting knowledge encoded in their activations.☆186Updated this week
- Erasing concepts from neural representations with provable guarantees☆209Updated last week
- This repository collects all relevant resources about interpretability in LLMs☆288Updated 2 weeks ago
- A library for efficient patching and automatic circuit discovery.☆31Updated last month
- ☆253Updated 8 months ago
- Notebooks accompanying Anthropic's "Toy Models of Superposition" paper☆97Updated 2 years ago
- Sparse Autoencoder Training Library☆27Updated 3 weeks ago
- LLM experiments done during SERI MATS - focusing on activation steering / interpreting activation spaces☆78Updated last year
- ☆170Updated 8 months ago
- A toolkit for describing model features and intervening on those features to steer behavior.☆99Updated last week
- Sparse probing paper full code.☆51Updated 11 months ago