Keep Apart Research Going: Donate Today
Mar 10, 2025
HalluShield: A Mechanistic Approach to Hallucination Resistant Models
Tanzeel Shaikh, Naiyarah Hussain, Samuel Nihoul, Sonal Joshi, Lucia de la Torre
Summary
Our project tackles the critical problem of hallucinations in large language models (LLMs) used in healthcare settings, where inaccurate information can have serious consequences. We developed a proof-of-concept system that classifies LLM-generated responses as either factual or hallucinated. Our approach leverages sparse autoencoders (GoodFire’s Ember) trained on neural activations from Meta Llama 3. These autoencoders identify monosemantic features that serve as strong indicators of hallucination patterns. By feeding these extracted features into tree-based classification models (XGBoost), we achieved an impressive F1 score of 89% on our test dataset. This machine learning approach offers several advantages over traditional methods and LLM as a judge. First, it can be specifically trained on in-domain datasets (eg: medical) for domain-specific hallucination detection. Second, the model is interpretable, showing which activation patterns correlate with hallucinations and acts as a post-processing layer applied to LLM output.
Cite this work:
@misc {
title={
HalluShield: A Mechanistic Approach to Hallucination Resistant Models
},
author={
Tanzeel Shaikh, Naiyarah Hussain, Samuel Nihoul, Sonal Joshi, Lucia de la Torre
},
date={
3/10/25
},
organization={Apart Research},
note={Research submission to the research sprint hosted by Apart.},
howpublished={https://apartresearch.com}
}
