Code and results accompanying the paper "Refusal in Language Models Is Mediated by a Single Direction".
☆390Jun 13, 2025Updated 11 months ago
Alternatives and similar repositories for refusal_direction
Users that are interested in refusal_direction are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- ☆18Mar 30, 2025Updated last year
- ☆25Jun 13, 2024Updated last year
- Steering Llama 2 with Contrastive Activation Addition☆231May 23, 2024Updated 2 years ago
- A fast + lightweight implementation of the GCG algorithm in PyTorch☆331May 13, 2025Updated last year
- Improving Alignment and Robustness with Circuit Breakers☆262Sep 24, 2024Updated last year
- Bare Metal GPUs on DigitalOcean Gradient AI • AdPurpose-built for serious AI teams training foundational models, running large-scale inference, and pushing the boundaries of what's possible.
- Fluent student-teacher redteaming☆23Jul 25, 2024Updated last year
- Official Repository for ACL 2024 Paper SafeDecoding: Defending against Jailbreak Attacks via Safety-Aware Decoding☆152Jul 19, 2024Updated last year
- Code repo of our paper Towards Understanding Jailbreak Attacks in LLMs: A Representation Space Analysis (https://arxiv.org/abs/2406.10794…☆24Jul 26, 2024Updated last year
- [ICLR 2025] Official Repository for "Tamper-Resistant Safeguards for Open-Weight LLMs"☆67Jun 9, 2025Updated 11 months ago
- Official Repository for The Paper: Safety Alignment Should Be Made More Than Just a Few Tokens Deep☆183Apr 23, 2025Updated last year
- Training Sparse Autoencoders on Language Models☆1,389Updated this week
- A resource repository for representation engineering in large language models☆153Nov 14, 2024Updated last year
- A library for mechanistic interpretability of GPT-style language models☆3,461Updated this week
- ☆416Aug 21, 2025Updated 9 months ago
- GPU virtual machines on DigitalOcean Gradient AI • AdGet to production fast with high-performance AMD and NVIDIA GPUs you can spin up in seconds. The definition of operational simplicity.
- Mechanistic Interpretability Visualizations using React☆348Apr 30, 2026Updated 3 weeks ago
- [ICML 2024] Assessing the Brittleness of Safety Alignment via Pruning and Low-Rank Modifications☆90Mar 30, 2025Updated last year
- Delphi was the home of a temple to Phoebus Apollo, which famously had the inscription, 'Know Thyself.' This library lets language models …☆256May 18, 2026Updated last week
- Sparse Autoencoder Training Library☆57May 1, 2025Updated last year
- [ICLR 2025] General-purpose activation steering library☆168Sep 18, 2025Updated 8 months ago
- Jailbreaking Leading Safety-Aligned LLMs with Simple Adaptive Attacks [ICLR 2025]☆386Jan 23, 2025Updated last year
- [ICLR 2025] BlueSuffix: Reinforced Blue Teaming for Vision-Language Models Against Jailbreak Attacks☆31Nov 2, 2025Updated 6 months ago
- Improved Few-Shot Jailbreaking Can Circumvent Aligned Language Models and Their Defenses (NeurIPS 2024)☆65Jan 11, 2025Updated last year
- The nnsight package enables interpreting and manipulating the internals of deep learned models.☆930May 18, 2026Updated last week
- Managed hosting for WordPress and PHP on Cloudways • AdManaged hosting for WordPress, Magento, Laravel, or PHP apps, on multiple cloud providers. Deploy in minutes on Cloudways by DigitalOcean.
- HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal☆953Aug 16, 2024Updated last year
- Official repository for ICML 2024 paper "On Prompt-Driven Safeguarding for Large Language Models"☆108May 20, 2025Updated last year
- ☆33Jun 24, 2024Updated last year
- ☆38Apr 30, 2024Updated 2 years ago
- [ICML 2025] Weak-to-Strong Jailbreaking on Large Language Models☆90May 2, 2025Updated last year
- ☆214Oct 14, 2025Updated 7 months ago
- Simple Python library/structure to ablate features in LLMs which are supported by TransformerLens☆642Jun 11, 2024Updated last year
- ☆208Nov 17, 2024Updated last year
- ☆202Nov 26, 2023Updated 2 years ago
- Managed Kubernetes at scale on DigitalOcean • AdDigitalOcean Kubernetes includes the control plane, bandwidth allowance, container registry, automatic updates, and more for free.
- ☆27Oct 6, 2024Updated last year
- LLM experiments done during SERI MATS - focusing on activation steering / interpreting activation spaces☆103Sep 21, 2023Updated 2 years ago
- [ICML 2024] COLD-Attack: Jailbreaking LLMs with Stealthiness and Controllability☆176Dec 18, 2024Updated last year
- Tools for understanding how transformer predictions are built layer-by-layer☆589Aug 7, 2025Updated 9 months ago
- Create feature-centric and prompt-centric visualizations for sparse autoencoders (like those from Anthropic's published research).☆259Feb 27, 2026Updated 2 months ago
- Multi-dimensional analysis of orthogonal safety directions in LLM alignment☆22Mar 20, 2025Updated last year
- [COLM 2024] JailBreakV-28K: A comprehensive benchmark designed to evaluate the transferability of LLM jailbreak attacks to MLLMs, and fur…☆93May 9, 2025Updated last year