EleutherAI / concept-erasure
Erasing concepts from neural representations with provable guarantees
☆208Updated 3 weeks ago
Related projects ⓘ
Alternatives and complementary repositories for concept-erasure
- Mechanistic Interpretability Visualizations using React☆195Updated 3 months ago
- Create feature-centric and prompt-centric visualizations for sparse autoencoders (like those from Anthropic's published research).☆157Updated last month
- ☆99Updated this week
- Tools for understanding how transformer predictions are built layer-by-layer☆429Updated 5 months ago
- Keeping language models honest by directly eliciting knowledge encoded in their activations.☆186Updated this week
- Extract full next-token probabilities via language model APIs☆228Updated 8 months ago
- ☆186Updated last month
- ☆108Updated last year
- ☆102Updated last month
- Steering vectors for transformer language models in Pytorch / Huggingface☆64Updated last month
- TART: A plug-and-play Transformer module for task-agnostic reasoning☆190Updated last year
- Sparse autoencoders☆333Updated 2 weeks ago
- Steering Llama 2 with Contrastive Activation Addition☆94Updated 5 months ago
- ☆99Updated 3 months ago
- ☆252Updated 8 months ago
- Code to reproduce "Transformers Can Do Arithmetic with the Right Embeddings", McLeish et al (NeurIPS 2024)☆176Updated 5 months ago
- Code and results accompanying the paper "Refusal in Language Models Is Mediated by a Single Direction".☆117Updated last month
- Understand and test language model architectures on synthetic tasks.☆161Updated 6 months ago
- Training Sparse Autoencoders on Language Models☆449Updated this week
- A toolkit for describing model features and intervening on those features to steer behavior.☆69Updated this week
- ☆73Updated last month
- RuLES: a benchmark for evaluating rule-following in language models☆210Updated last month
- Emergent world representations: Exploring a sequence model trained on a synthetic task☆168Updated last year
- ☆24Updated 7 months ago
- ☆43Updated 4 months ago
- git extension for {collaborative, communal, continual} model development☆205Updated 5 months ago
- Code repository for the c-BTM paper☆105Updated last year
- Datasets collection and preprocessings framework for NLP extreme multitask learning☆147Updated 3 months ago
- ☆239Updated 4 months ago
- ☆73Updated 4 months ago