andyrdt / refusal_directionLinks
Code and results accompanying the paper "Refusal in Language Models Is Mediated by a Single Direction".
☆223Updated 8 months ago
Alternatives and similar repositories for refusal_direction
Users that are interested in refusal_direction are comparing it to the libraries listed below
Sorting:
- Steering vectors for transformer language models in Pytorch / Huggingface☆101Updated 3 months ago
- Improving Alignment and Robustness with Circuit Breakers☆208Updated 8 months ago
- Steering Llama 2 with Contrastive Activation Addition☆154Updated 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
- Röttger et al. (NAACL 2024): "XSTest: A Test Suite for Identifying Exaggerated Safety Behaviours in Large Language Models"☆97Updated 3 months ago
- LLM experiments done during SERI MATS - focusing on activation steering / interpreting activation spaces☆93Updated last year
- Function Vectors in Large Language Models (ICLR 2024)☆167Updated last month
- Using sparse coding to find distributed representations used by neural networks.☆247Updated last year
- General-purpose activation steering library☆74Updated 3 weeks ago
- A toolkit for describing model features and intervening on those features to steer behavior.☆184Updated 6 months ago
- For OpenMOSS Mechanistic Interpretability Team's Sparse Autoencoder (SAE) research.☆116Updated this week
- Create feature-centric and prompt-centric visualizations for sparse autoencoders (like those from Anthropic's published research).☆200Updated 5 months ago
- ☆120Updated 6 months ago
- Sparse Autoencoder for Mechanistic Interpretability☆248Updated 10 months ago
- WMDP is a LLM proxy benchmark for hazardous knowledge in bio, cyber, and chemical security. We also release code for RMU, an unlearning m…☆121Updated this week
- Benchmarking LLMs with Challenging Tasks from Real Users☆223Updated 7 months ago
- ☆170Updated last month
- Algebraic value editing in pretrained language models☆65Updated last year
- Repo accompanying our paper "Do Llamas Work in English? On the Latent Language of Multilingual Transformers".☆76Updated last year
- ☆174Updated last month
- Stanford NLP Python library for benchmarking the utility of LLM interpretability methods☆89Updated last week
- Code for In-context Vectors: Making In Context Learning More Effective and Controllable Through Latent Space Steering☆174Updated 3 months ago
- ☆209Updated last year
- AI Logging for Interpretability and Explainability🔬☆119Updated 11 months ago
- Sparsify transformers with SAEs and transcoders☆547Updated last week
- Open One-Stop Moderation Tools for Safety Risks, Jailbreaks, and Refusals of LLMs☆77Updated 6 months ago
- We jailbreak GPT-3.5 Turbo’s safety guardrails by fine-tuning it on only 10 adversarially designed examples, at a cost of less than $0.20…☆298Updated last year
- ☆116Updated 9 months ago
- ☆83Updated 9 months ago
- ☆97Updated last month