Emily J. Herron

Postdoctoral Research Associate at Oak Ridge National Laboratory
Analytics and AI Methods at Scale Group
emily.j.herron@gmail.com | CV | Google Scholar | LinkedIn | Github

About

Emily Herron is a postdoctoral research associate in the Analytics & AI Methods at Scale Group in the National Center for Computational Science at Oak Ridge National Laboratory. Her current research focuses on machine learning, particularly in neural architecture search, large language models, and AI trustworthiness. She completed her Ph.D. in Data Science & Engineering from the University of Tennessee, where she worked on generalized differentiable neural architecture search with scaling and stability improvements.

Current Work

• Developing pipelines for assessing LLM trustworthiness in scientific applications, focusing on truthfulness, adversarial robustness, and ethics
• Incorporating evolutionary neural architecture search into mixture of experts (MoE) transformer architectures at scale
• Researching LLM-based hypothesis generation in fields of natural language processing and materials science

Publications

ChatHPC: Empowering HPC Users with Large Language Models
J. Yin et al.
Journal of Supercomputing, vol. 81, 2025
DOI: https://doi.org/10.1007/s11227-024-06637-1

Exploring Scientific Hypothesis Generation with Mamba
M. Chai, E. Herron, E. Cervantes, and T. Ghosal
Proceedings of the 1st Workshop on NLP for Science (NLP4Science), ACL, 2024
Pages 197-207

SciTrust: Evaluating the Trustworthiness of Large Language Models for Science
E. Herron, J. Yin, and F. Wang
AI4S: 5th Workshop on Artificial Intelligence and Machine Learning for Scientific Applications, 2024

ICDARTS: Improving the Stability and Performance of Cyclic DARTS
E. Herron, D. Rose, and S. Young
arXiv, September 2023
https://arxiv.org/abs/2309.00664

ICDARTS: Improving the Stability of Cyclic DARTS
E. J. Herron, S. R. Young, and D. Rose
2022 21st IEEE International Conference on Machine Learning and Applications (ICMLA), 2022

The Sensitivity of Word Embeddings-Based Author Detection Models to Semantic-Preserving Adversarial Perturbations
J. Duncan et al.
arXiv, 2021
DOI: 10.48550/ARXIV.2102.11917
https://arxiv.org/abs/2102.11917

Ensembles of Networks Produced from Neural Architecture Search
E. J. Herron, S. R. Young, and T. E. Potok
International Conference on High Performance Computing, 2020
Pages 223-234

Applying Image Feature Extraction to Cluttered Scientific Repositories
E. Herron, T. J. Skluzacek, I. Foster, and K. Chard
2017

Selected Presentations

SciTrust: Evaluating the Trustworthiness of Large Language Models for Science
2024 Monterey Data Conference, Monterey, California
Poster Presentation

Generalized Differentiable Neural Architecture Search with Performance and Stability Improvements for Scientific Applications
SOS26 2024, Cocoa Beach, FL
Poster Presentation

ICDARTS: Improving the Stability of Cyclic DARTS
2022 21st IEEE International Conference on Machine Learning and Applications
Nassau, The Bahamas

Ensembles of Neural Networks Produced from Neural Architecture Search
Women in High Performance Computing Workshop, SuperComputing 2020
Virtual Presentation

Ensembles of Neural Networks Produced from Neural Architecture Search
The International Conference on High Performance Computing 2020
Virtual Presentation

Applying Image Feature Extraction to Cluttered Scientific Repositories
Student Research Competition Poster Session, SuperComputing 2017
Denver, CO

Education

Ph.D. in Data Science & Engineering
University of Tennessee, Knoxville
2018 - 2023

• Thesis: Generalized Differentiable Neural Architecture Search with Scaling and Stability Improvements
• Advisor: Dr. Steven R. Young
• GPA: 3.95 / 4.00

B.S. in Computational Science
Mercer University
2014 - 2018

• Graduated Summa Cum Laude
• GPA: 3.94 / 4.00

Experience

Oak Ridge National Laboratory
Postdoctoral Research Associate (Jan 2024 - Present)

• Developing pipeline for assessing the trustworthiness of large language models for science
• Incorporating evolutionary neural architecture search into mixture of experts (MoE) transformer architectures
• Researching LLM-based hypothesis generation in NLP and materials science

Graduate Research Assistant (Aug 2018 - Dec 2023)

• Developed stability improvements to CDARTS NAS algorithm
• Researched and implemented selection algorithms for ORNL’s MENNDL NAS software
• Collaborated on creation of NIEHS document mining pipeline

Professional Service

• ORNL Traveling Science Fair Volunteer (2024)
• ORNL Undergraduate Summer Internship Program Mentor (2024)
• Assistant Instructor, ORNLAI Summer Institute Tutorial (2024)
• ICML 2022 Reviewer - Top 10%
• Bredesen Center Peer Mentor (2022-2023)

Awards & Honors

• OLCF Director’s Discretion Project - 20,000 Summit Hours (2022-2023)
• Bredesen Center Data Science & Engineering Fellowship (2018)
• ACM Student Poster Competition Undergraduate Semifinalist (2017)
• Outstanding Student in Computational Science, Mercer University (2016-2018)
• Summa Cum Laude, Mercer University (2018)

Professional Affiliations

• Institute of Electrical and Electronics Engineers (IEEE) Member (2022 - Present)
• Association for Computing Machinery (ACM) Member (2017 - Present)