---
title: README
emoji: 🐢
colorFrom: green
colorTo: blue
sdk: static
pinned: false
---

# PhillipsLab — University of Utah

<img src="kahlert_logo.png" alt="Kahlert School of Computing" width="200">


## About Us
**PhillipsLab** is part of the Kahlert School of Computing at the University of Utah, led by Prof. Jeff M. Phillips.

We study **interpretable AI**, **activation steering** and **dataset curation**.  
Our work bridges theory and application — from understanding internal representations in large language models to make models more truthful and safe .

## Research Focus
- Representation and activation steering in large language models  
- Dataset design for interpretable and steerable AI  
  

## Purpose of this Hub
This Hugging Face organization hosts:
- Private and public **datasets** for steering and interpretability research  
- Experiment **models** and **evaluation code**  
- Collaborative resources for our research group  

## Contact
**Prof. Jeff M. Phillips**    
🌐 [users.cs.utah.edu/~jeffp](https://users.cs.utah.edu/~jeffp/)

---

*PhillipsLab © 2025 — University of Utah*