Selected Publications

Full spectrum modeling of in situ gamma-ray detector measurements with a focus on precipitation-induced transients

Published in Journal of Environmental Radioactivity, 2026

We analyze and model gamma-ray spectra measured during rainfall by a NaI(Tl) detector to derive accurate source terms for Pb-214 and Bi-214 at high frequency. This is the first time full-spectrum modeling has been used for continuous measurements of radon progeny at rates of more than once per minute.

Recommended citation: Bandstra, M.S., Ghawaly, J.M., Peplow, D.E., Archer, D.E., Quiter, B.J., Joshi, T.H.Y., Nicholson, A.D., Willis, M.J., Garishvili, I., Rowe, A.J., Longmire, B.R., & Nattress, J.T. (2026). Full spectrum modeling of in situ gamma-ray detector measurements with a focus on precipitation-induced transients. Journal of Environmental Radioactivity, 291, 107826. https://doi.org/10.1016/j.jenvrad.2025.107826

ForensicLLM: A Local Large Language Model for Digital Forensics

Published in 2025 DFRWS EU, 2025

In this work, we introduce ForensicLLM, a 4-bit quantized LLaMA-3.1-8B model fine-tuned on Q&A samples extracted from digital forensic research articles and curated digital artifacts. We evaluate the model’s performance, both quantitatively and qualitatively, against standard RAG and base-model performance.

Download here

REx86: A Local Large Language Model for Assisting in x86 Assembly Reverse Engineering

Published in Annual Computer Security Applications Conference (ACSAC), 2025

We evaluate parameter-efficient fine-tuned local LLMs for assisting with x86 reverse engineering tasks. Eight open-weight models are fine-tuned on a curated dataset of 5,981 x86 assembly examples, with the fine-tuned Qwen2.5-Coder-7B (REx86) emerging as the top performer.

Recommended citation: Lea, D., Ghawaly, J., Richard III, G., Ali-Gombe, A., & Case, A. (2025). REx86: A Local Large Language Model for Assisting in x86 Assembly Reverse Engineering. Annual Computer Security Applications Conference (ACSAC). https://jamesghawaly.org/files/ACSAC25_x86RELLM_cp.pdf

Exploring Spiking Neural Networks for Binary Classification in Multivariate Time Series at the Edge

Published in IJCNN, 2025

We present a general framework for training spiking neural networks (SNNs) to perform binary classification on multivariate time series, with a focus on step-wise prediction and high precision at low false alarm rates.

Recommended citation: Ghawaly, J., Nicholson, A., Schuman, C., Diez, D., Young, A., & Witherspoon, B. (2025). Exploring Spiking Neural Networks for Binary Classification in Multivariate Time Series at the Edge. 2025 International Joint Conference on Neural Networks (IJCNN). In press.

Exploring Large Language Models for Semantic Analysis and Categorization of Android Malware

Published in 2024 Annual Computer Security Applications Conference Workshops (ACSAC Workshops), 2024

In this paper, we explore leveraging Large Language Models (LLMs) for semantic malware analysis to expedite the analysis of known and novel samples. Built on GPT-4o-mini model, MalParse is designed to augment malware analysis for Android through a hierarchical-tiered summarization chain and strategic prompt engineering.

Download here

Performance Optimization Study of the Neuromorphic Radiation Anomaly Detector

Published in 2023 International Conference on Neuromorphic Systems, 2023

This work reports on new results and insights from the optimization of spiking neural networks developed for gamma-ray radiation anomaly detection.

Explaining machine-learning models for gamma-ray detection and identification

Published in PLoS ONE, 2023

This work evaluates a variety of neural network explainability approaches for explaining deep learning models trained to perform gamma-ray spectroscopy.

A Neuromorphic Algorithm for Radiation Anomaly Detection

Published in 2022 International Conference on Neuromorphic Systems, 2022

In this work, we present initial results on the development of a neuromorphic spiking neural network for performing gamma-ray radiation anomaly detection, the first known application of neuromorphic computing to be applied to the radiation detection domain.

TNM Tumor Classification from Unstructured Breast Cancer Pathology Reports using LoRA Finetuning of Mistral 7B

Published in AAAI 2024 Spring Symposium on Clinical Foundation Models, 2022

In this paper, we explore the application of Low-Rank Adaptation (LoRA) fine-tuning of small language models for performing TNM staging on unstructured pathology reports for triple negative breast cancer cases. We also attempt to develop a more generalized approach, so that our work can be applied to other NLP tasks within the medical field.

Characterization of the Autoencoder Radiation Anomaly Detection (ARAD) model

Published in Engineering Applications of Artificial Intelligence, 2022

In this work we demonstrate an in-depth analysis and characterization of the Autoencoder Radiation Anomaly Detection (ARAD) algorithm. ARAD is a deep convolutional autoencoder designed to detect anomalous radioactive signatures in gamma-ray spectra collected by NaI(Tl) detectors.

Data for Training and Testing Radiation Detection Algorithms in an Urban Environment

Published in Nature Scientific Data, 2020

To encourage the development of new detection, radioisotope identification, and source localization algorithms, a dataset consisting of realistic Monte Carlo–simulated radiation detection data from a 2 in. × 4 in. × 16 in. NaI(Tl) scintillation detector moving through a simulated urban environment based on Knoxville, Tennessee, was developed and made public in the form of a Topcoder competition.

Download here