nrimsky / LM-exp
LLM experiments done during SERI MATS - focusing on activation steering / interpreting activation spaces
☆91Updated last year
Alternatives and similar repositories for LM-exp:
Users that are interested in LM-exp are comparing it to the libraries listed below
- Steering Llama 2 with Contrastive Activation Addition☆126Updated 9 months ago
- A Mechanistic Understanding of Alignment Algorithms: A Case Study on DPO and Toxicity.☆69Updated last week
- Röttger et al. (NAACL 2024): "XSTest: A Test Suite for Identifying Exaggerated Safety Behaviours in Large Language Models"☆86Updated 3 weeks ago
- A resource repository for representation engineering in large language models☆111Updated 4 months ago
- For OpenMOSS Mechanistic Interpretability Team's Sparse Autoencoder (SAE) research.☆95Updated this week
- Steering vectors for transformer language models in Pytorch / Huggingface☆90Updated 3 weeks ago
- Algebraic value editing in pretrained language models☆63Updated last year
- A library for efficient patching and automatic circuit discovery.☆57Updated last month
- ☆210Updated 5 months ago
- ☆80Updated 7 months ago
- Using sparse coding to find distributed representations used by neural networks.☆224Updated last year
- ☆193Updated last year
- General-purpose activation steering library☆50Updated 2 months ago
- ☆113Updated 7 months ago
- Delphi was the home of a temple to Phoebus Apollo, which famously had the inscription, 'Know Thyself.' This library lets language models …☆161Updated this week
- Repo accompanying our paper "Do Llamas Work in English? On the Latent Language of Multilingual Transformers".☆70Updated last year
- datasets from the paper "Towards Understanding Sycophancy in Language Models"☆73Updated last year
- ☆38Updated last year
- Improving Alignment and Robustness with Circuit Breakers☆190Updated 5 months ago
- ☆149Updated last week
- ☆92Updated last year
- Code and results accompanying the paper "Refusal in Language Models Is Mediated by a Single Direction".☆192Updated 5 months ago
- ☆22Updated 3 weeks ago
- ☆53Updated 2 years ago
- Code for my NeurIPS 2024 ATTRIB paper titled "Attribution Patching Outperforms Automated Circuit Discovery"☆29Updated 9 months ago
- Code release for "Debating with More Persuasive LLMs Leads to More Truthful Answers"☆103Updated 11 months ago
- Function Vectors in Large Language Models (ICLR 2024)☆144Updated 5 months ago
- ☆68Updated this week
- Sparse probing paper full code.☆55Updated last year
- Codes and datasets of the paper Red-Teaming Large Language Models using Chain of Utterances for Safety-Alignment☆95Updated last year