A complete end-to-end pipeline for LLM interpretability with sparse autoencoders (SAEs) using Llama 3.2, written in pure PyTorch and fully reproducible.
This project explores methods to detect and mitigate jailbreak behaviors in Large Language Models (LLMs). By analyzing activation patterns—particularly in deeper layers—we identify distinct differences between compliant and non-compliant responses to uncover a jailbreak "direction." Using this insight, we develop intervention strategies that modify
Automates attribution-graph analysis via probe prompting: circuit-trace a prompt, auto-generate concept probes, profile feature activations, cluster supernodes.
🔬 HexRay: An Open-Source Neuroscope for AI — Tracing Tokens, Neurons, and Decisions for Frontier AI Research, Safety, and Security
Add a description, image, and links to the llm-interpretability topic page so that developers can more easily learn about it.
To associate your repository with the llm-interpretability topic, visit your repo's landing page and select "manage topics."