Eric Michael Smith
Research scientist at Meta; first author of HolisticBias and a core contributor to Llama 3
- Research Scientist (the page's bio describes a Research Engineer role at FAIR), Meta (Facebook AI Research, FAIR)19
Eric Michael Smith is a research scientist at Meta; Meta's people page describes Smith as a research engineer at Facebook AI Research (FAIR). Smith is first author of the HolisticBias paper (2022), which presents a bias-measurement dataset of nearly 600 descriptor terms across 13 demographic axes, and co-authored Llama 2 (2023), Llama Guard 3 Vision (2024), and the open-domain chatbot papers Recipes for Building an Open-Domain Chatbot (2020) and BlenderBot 3 (2022). Smith is listed among the core contributors of The Llama 3 Herd of Models (2024) and among the authors of Meta's Muse Spark Safety & Preparedness Report (2026).12467859
Work
Llama 2, Llama 3, and Llama Guard 3 VisionOpen weights
Co-author of Llama 2 (2023), listed among the core contributors of The Llama 3 Herd of Models (2024), and co-author of Llama Guard 3 Vision (2024), a multimodal safeguard model fine-tuned on Llama 3.2-Vision.4561
Muse Spark Safety & Preparedness ReportPublished research
Listed among the authors of the Muse Spark Safety & Preparedness Report (2026), which presents Meta's catastrophic-risk evaluations and broader safety analysis for the Muse Spark model.9
HolisticBias (external site: github.com)Open-source code
First author of "I'm sorry to hear that": Finding New Biases in Language Models with a Holistic Descriptor Dataset (2022), which introduces HolisticBias; the facebookresearch/ResponsibleNLP license file publishes the HolisticBias code under the MIT License.23
This profile covers public professional work only, from the sources below. It is not an endorsement or a ranking. To correct or remove it, use the correction form.