montemac / activation_additions
Algebraic value editing in pretrained language models
☆54Updated 10 months ago
Related projects: ⓘ
- Steering Llama 2 with Contrastive Activation Addition☆83Updated 3 months ago
- ☆75Updated this week
- Sparse probing paper full code.☆47Updated 9 months ago
- LLM experiments done during SERI MATS - focusing on activation steering / interpreting activation spaces☆73Updated 11 months ago
- ☆91Updated last month
- ☆69Updated 10 months ago
- A Mechanistic Understanding of Alignment Algorithms: A Case Study on DPO and Toxicity.☆46Updated last month
- Function Vectors in Large Language Models (ICLR 2024)☆107Updated last month
- ☆64Updated last month
- Code for my NeurIPS 2024 ATTRIB paper titled "Attribution Patching Outperforms Automated Circuit Discovery"☆22Updated 3 months ago
- Using sparse coding to find distributed representations used by neural networks.☆162Updated 10 months ago
- A library for efficient patching and automatic circuit discovery.☆18Updated 3 weeks ago
- Code and results accompanying the paper "Refusal in Language Models Is Mediated by a Single Direction".☆76Updated 3 weeks ago
- Code for the ICLR 2024 paper "How to catch an AI liar: Lie detection in black-box LLMs by asking unrelated questions"☆56Updated 3 months ago
- Code release for "Debating with More Persuasive LLMs Leads to More Truthful Answers"☆70Updated 5 months ago
- ☆159Updated 6 months ago
- ☆174Updated 4 months ago
- ☆68Updated 7 months ago
- Röttger et al. (2023): "XSTest: A Test Suite for Identifying Exaggerated Safety Behaviours in Large Language Models"☆55Updated 8 months ago
- A resource repository for representation engineering in large language models☆36Updated last week
- The accompanying code for "Transformer Feed-Forward Layers Are Key-Value Memories". Mor Geva, Roei Schuster, Jonathan Berant, and Omer Le…☆80Updated 3 years ago
- Steering vectors for transformer language models in Pytorch / Huggingface☆52Updated last month
- Inspecting and Editing Knowledge Representations in Language Models☆105Updated last year
- ☆33Updated 3 months ago
- datasets from the paper "Towards Understanding Sycophancy in Language Models"☆59Updated 10 months ago
- ☆54Updated last week
- ☆24Updated 4 months ago
- AI Logging for Interpretability and Explainability🔬☆74Updated 3 months ago
- [NeurIPS'23] Aging with GRACE: Lifelong Model Editing with Discrete Key-Value Adaptors☆64Updated 6 months ago
- ☆74Updated this week