EleutherAI / concept-erasureLinks
Erasing concepts from neural representations with provable guarantees
☆227Updated 4 months ago
Alternatives and similar repositories for concept-erasure
Users that are interested in concept-erasure are comparing it to the libraries listed below
Sorting:
- Mechanistic Interpretability Visualizations using React☆251Updated 5 months ago
- ☆121Updated last year
- Create feature-centric and prompt-centric visualizations for sparse autoencoders (like those from Anthropic's published research).☆200Updated 5 months ago
- ☆223Updated 8 months ago
- Keeping language models honest by directly eliciting knowledge encoded in their activations.☆204Updated last week
- Steering vectors for transformer language models in Pytorch / Huggingface☆101Updated 3 months ago
- ☆269Updated last year
- Extract full next-token probabilities via language model APIs☆248Updated last year
- Delphi was the home of a temple to Phoebus Apollo, which famously had the inscription, 'Know Thyself.' This library lets language models …☆180Updated this week
- ☆96Updated 3 months ago
- ☆28Updated last year
- Code to reproduce "Transformers Can Do Arithmetic with the Right Embeddings", McLeish et al (NeurIPS 2024)☆190Updated last year
- Utilities for the HuggingFace transformers library☆68Updated 2 years ago
- Tools for understanding how transformer predictions are built layer-by-layer☆493Updated last year
- A library for efficient patching and automatic circuit discovery.☆65Updated last month
- ☆131Updated 2 months ago
- ☆97Updated last month
- Sparse Autoencoder Training Library☆50Updated last month
- Steering Llama 2 with Contrastive Activation Addition☆154Updated last year
- Notebooks accompanying Anthropic's "Toy Models of Superposition" paper☆126Updated 2 years ago
- Mechanistic Interpretability for Transformer Models☆51Updated 3 years ago
- Sparsify transformers with SAEs and transcoders☆547Updated last week
- A toolkit for describing model features and intervening on those features to steer behavior.☆184Updated 6 months ago
- ☆93Updated 3 months ago
- Comprehensive analysis of difference in performance of QLora, Lora, and Full Finetunes.☆81Updated last year
- Code and results accompanying the paper "Refusal in Language Models Is Mediated by a Single Direction".☆223Updated 8 months ago
- The simplest, fastest repository for training/finetuning medium-sized GPTs.☆126Updated 3 weeks ago
- MiniHF is an inference, human preference data collection, and fine-tuning tool for local language models. It is intended to help the user…☆172Updated last week
- Code for reproducing our paper "Not All Language Model Features Are Linear"☆75Updated 6 months ago
- Emergent world representations: Exploring a sequence model trained on a synthetic task☆181Updated last year