Code to enable layer-level steering in LLMs using sparse auto encoders
☆35Sep 18, 2025Updated 11 months ago
Alternatives and similar repositories for sae-steering
Users that are interested in sae-steering are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- Improving Steering Vectors by Targeting Sparse Autoencoder Features☆30Nov 20, 2024Updated last year
- [ICLR 2025] General-purpose activation steering library☆185Sep 18, 2025Updated 11 months ago
- [ICML 2025 Poster] SAE-V: Interpreting Multimodal Models for Enhanced Alignment☆18Jun 5, 2025Updated last year
- Materials for "Multi-property Steering of Large Language Models with Dynamic Activation Composition"☆14Nov 22, 2024Updated last year
- [NAACL'25 Oral] Steering Knowledge Selection Behaviours in LLMs via SAE-Based Representation Engineering☆85Jun 20, 2026Updated 2 months ago
- Deploy to Railway using AI coding agents - Free Credits Offer • AdUse Claude Code, Codex, OpenCode, and more. Autonomous software development now has the infrastructure to match with Railway.
- A TinyStories LM with SAEs and transcoders☆16Aug 19, 2026Updated 2 weeks ago
- PyTorch and NNsight implementation of AtP* (Kramar et al 2024, DeepMind)☆21Jan 19, 2025Updated last year
- Official Code for our paper: "Language Models Learn to Mislead Humans via RLHF""☆20Oct 11, 2024Updated last year
- Archive of questions from the Cambridge Mathematics Tripos☆10Jun 6, 2022Updated 4 years ago
- Code for the "Overcoming Sparsity Artifacts in Crosscoders to Interpret Chat-Tuning" paper.☆18Jul 6, 2026Updated last month
- [ICML 2025] Unlearning in Diffusion Models using Sparse Autoencoders☆63Oct 16, 2025Updated 10 months ago
- ☆13Dec 4, 2024Updated last year
- Code for my NeurIPS 2024 ATTRIB paper titled "Attribution Patching Outperforms Automated Circuit Discovery"☆49May 31, 2024Updated 2 years ago
- Performant framework for training, analyzing and visualizing Sparse Autoencoders (SAEs) and their frontier variants.☆227Updated this week
- Deploy on Railway without the complexity - Free Credits Offer • AdConnect your repo and Railway handles the rest with instant previews. Quickly provision container image services, databases, and storage volumes.
- [ICLR26] Safety Mirage: How Spurious Correlations Undermine VLM Safety Fine-Tuning and Can Be Mitigated by Machine Unlearning☆21Apr 16, 2026Updated 4 months ago
- A framework for implementing equivariant DL☆10May 25, 2021Updated 5 years ago
- Official repo of "Towards Interpretable Protein Structure Prediction with Sparse Autoencoders" published at ICLR 2025 GEM workshop.☆17Mar 13, 2025Updated last year
- ☆15Jan 2, 2026Updated 7 months ago
- Learning Safety Constraints for Large Language Models (ICML2025)☆35May 25, 2026Updated 3 months ago
- ☆74Jan 17, 2025Updated last year
- Implementation of the BatchTopK activation function for training sparse autoencoders (SAEs)☆67Jul 24, 2025Updated last year
- A library for efficient patching and automatic circuit discovery.☆101Aug 17, 2026Updated 2 weeks ago
- Steering Llama 2 with Contrastive Activation Addition☆250May 23, 2024Updated 2 years 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.
- ☆601Jul 19, 2024Updated 2 years ago
- ☆13Oct 28, 2022Updated 3 years ago
- A tiny easily hackable implementation of a feature dashboard.☆17Oct 21, 2025Updated 10 months ago
- Sparsify transformers with SAEs and transcoders☆739Aug 24, 2026Updated last week
- Code for "Preference Tuning For Toxicity Mitigation Generalizes Across Languages." Paper accepted at Findings of EMNLP 2024☆18Mar 25, 2025Updated last year
- Training Sparse Autoencoders on Language Models☆1,518Updated this week
- ☆16Feb 24, 2022Updated 4 years ago
- Optimizing diffusion for production-ready speeds☆40Jan 10, 2026Updated 7 months ago
- ☆26Jan 7, 2026Updated 7 months ago
- AI Agents on DigitalOcean Gradient AI Platform • AdBuild production-ready AI agents using customizable tools or access multiple LLMs through a single endpoint. Create custom knowledge bases or connect external data.
- HSTGODE code☆11Nov 26, 2023Updated 2 years ago
- Code repository for "Eliciting Secret Knowledge from Language Models"☆24Mar 30, 2026Updated 5 months ago
- 🪄 Interpreto is an interpretability toolbox for LLMs☆198Updated this week
- Extract residual-stream activations and apply steering vectors (including activation oracles) to any vLLM model during inference.☆121Aug 11, 2026Updated 3 weeks ago
- ☆36Jun 13, 2025Updated last year
- ☆10Nov 28, 2023Updated 2 years ago
- Topic supervised non-negative matrix factorization with sparse matrices☆12Mar 24, 2020Updated 6 years ago