HugoFry / mats_sae_training_for_ViTsLinks
☆21Updated last year
Alternatives and similar repositories for mats_sae_training_for_ViTs
Users that are interested in mats_sae_training_for_ViTs are comparing it to the libraries listed below
Sorting:
- ☆54Updated 10 months ago
- ☆127Updated last year
- ☆106Updated 7 months ago
- This is the official repository for the "Towards Vision-Language Mechanistic Interpretability: A Causal Tracing Tool for BLIP" paper acce…☆23Updated last year
- ☆187Updated 2 months ago
- ☆125Updated 2 weeks ago
- Modified to support crosscoder training.☆23Updated 2 months ago
- Improving Steering Vectors by Targeting Sparse Autoencoder Features☆24Updated 10 months ago
- Implementation of the BatchTopK activation function for training sparse autoencoders (SAEs)☆48Updated 2 months ago
- Open source replication of Anthropic's Crosscoders for Model Diffing☆59Updated 11 months ago
- Using sparse coding to find distributed representations used by neural networks.☆272Updated last year
- A library for efficient patching and automatic circuit discovery.☆77Updated 2 months ago
- Sparse Autoencoder Training Library☆54Updated 5 months ago
- Delphi was the home of a temple to Phoebus Apollo, which famously had the inscription, 'Know Thyself.' This library lets language models …☆215Updated this week
- Steering Llama 2 with Contrastive Activation Addition☆187Updated last year
- ☆242Updated last year
- ☆347Updated last month
- ☆174Updated 10 months ago
- Code for my NeurIPS 2024 ATTRIB paper titled "Attribution Patching Outperforms Automated Circuit Discovery"☆41Updated last year
- Create feature-centric and prompt-centric visualizations for sparse autoencoders (like those from Anthropic's published research).☆221Updated 9 months ago
- ☆18Updated last month
- Sparse Autoencoder for Mechanistic Interpretability☆268Updated last year
- ☆27Updated 2 years ago
- ☆23Updated 10 months ago
- ☆124Updated last year
- ☆71Updated last month
- ☆91Updated last year
- ☆23Updated 8 months ago
- Investigating the generalization behavior of LM probes trained to predict truth labels: (1) from one annotator to another, and (2) from e…☆28Updated last year
- ☆48Updated last year