Research
I study how neural networks, and language models in particular, learn and reason. I also put that understanding to work on real problems, from helping keep endangered languages alive to detecting tornadoes.
I lead the Language and Artificial Intelligence Research (LAIR) group at Tennessee Tech University.
Deep learning fundamentals
What can neural networks compute, why do they train, and how can we tell what they've actually learned? This work ranges from proofs about the computational power of decoder-only transformers to tools like Torch Canvas that make architectures, initialization and signal propagation visible.
Selected papers
- How Powerful are Decoder-Only Transformer Neural Models? IJCNN 2023 · cited by 68
- RL+ Transformer= A General-Purpose Problem Solver REALM Workshop 2025 · cited by 3
Torch CanvasInteractive demosDeep Learning course (AI 4200/5200)
6 papers 77 citations
Most cited in this area
- How Powerful are Decoder-Only Transformer Neural Models?
- Exploitation Is All You Need… for Exploration
- RL+ Transformer= A General-Purpose Problem Solver
- A Theoretical & Empirical Analysis of Transformer Language Model Behavior
- When Prediction Isn’t Enough: Decision-Criteria-Informed Learning for Threshold-Based Decisions
Large language models
How large language models represent uncertainty, reason, and learn, and how their behavior compares with human cognition. Topics include calibration and human-aligned uncertainty, memory effects, test-taking strategies, and alignment from feedback.
Selected papers
- The base-rate effect on llm benchmark performance: Disambiguating test-taking strategies from benchmark performance EMNLP Findings 2024 · cited by 10
- Using artificial populations to study psychological phenomena in neural models AAAI 2024 · cited by 10
- Large language model recall uncertainty is modulated by the fan effect CoNLL 2024 · cited by 9
- Chain of thought still thinks fast: Apricot helps with thinking slow CogSci 2024 · cited by 14
- Investigating Human-Aligned Large Language Model Uncertainty FLAIRS 2026 · cited by 3
CSC 7970 seminar on LLMsLLM foundations reading list
28 papers 191 citations
Most cited in this area
- How Powerful are Decoder-Only Transformer Neural Models?
- Leveraging large language models and machine learning for smart contract vulnerability detection
- Chain of thought still thinks fast: Apricot helps with thinking slow
- Investigating Expert-in-the-Loop LLM Discourse Patterns for Ancient Intertextual Analysis
- Do Large Language Models Learn Human-Like Strategic Preferences?
Games and decision making
How AI agents make decisions, on their own and when other players' choices matter. This work ranges from equilibria in classic games like the Traveler's Dilemma and models trained for threshold-based decisions to the strategic preferences of language models, how they cooperate as players in multiplayer games, and transformers that learn to act as general-purpose problem solvers.
Selected papers
- Do Large Language Models Learn Human-Like Strategic Preferences? REALM Workshop 2025 · cited by 11
- Exploitation Is All You Need… for Exploration IJCNN 2026 · cited by 5
- RL+ Transformer= A General-Purpose Problem Solver REALM Workshop 2025 · cited by 3
- Finding an Equilibrium in the Traveler's Dilemma with Fuzzy Weak Domination IEEE CoG 2021 · cited by 1
- When Prediction Isn’t Enough: Decision-Criteria-Informed Learning for Threshold-Based Decisions ICMLA 2026
6 papers 20 citations
Most cited in this area
- Do Large Language Models Learn Human-Like Strategic Preferences?
- Exploitation Is All You Need… for Exploration
- RL+ Transformer= A General-Purpose Problem Solver
- Finding an Equilibrium in the Traveler's Dilemma with Fuzzy Weak Domination
- When Prediction Isn’t Enough: Decision-Criteria-Informed Learning for Threshold-Based Decisions
Cherokee language NLP
Language technology in support of Cherokee (Tsalagi) revitalization, built with and for the community. Drawing on what worked for Irish (Gaeilge), the goal is a living resource that supports future generations of Cherokee speakers, along with the methods low-resource, polysynthetic languages need, such as better tokenization.
Selected for a small cohort that will shape what responsible, community-centered AI looks like in Indigenous language revitalization, collaborating on concrete projects over the 2027 fellowship year.
Selected papers
1 paper
Most cited in this area
In the news
- Tennessee Tech professor uses AI to help preserve Cherokee language Tennessee Tech News
- This Tennessee Tech professor is working to help preserve the Cherokee language using AI NewsChannel 5 Nashville
- Tennessee Tech uses AI to help preserve the Cherokee language WBIR 10 Knoxville
- Linguistic preservation Herald-Citizen
- Cherokee preservation with AI The Ledger (Knoxville)
Spiking neural networks Emerging
We are studying the fundamentals of spiking neural networks and their use in real-time environments. We are developing capability in this space now and have multiple papers in the works.
Early work
- Biologically Extending the Gen 2 ANN Model. FLAIRS 2019
1 paper
Most cited in this area
Tornado detection Emerging
We are working on tornado detection in areas where radar sweeps pass well above the ideal range for detecting couplet rotation, as they do across a large portion of Tennessee.
New work is in progress, and papers will appear here as they're published.
Funding
National Science Foundation · Award 2347337
ERI: Data-Driven Expansion of Electrified Transportation Network
National Science Foundation · Award 2526362
TN-THRIVE: Tennessee Health Research and Innovation using Verified AI Ecosystems
Collaborative Research: Conference: NAIRR Pilot Expansion: FA1: Mid-South Conferences on CyberInfrastructure Advances to Enable Interdisciplinary AI Research in Health
Collaborative award with the University of Tennessee, Knoxville, Meharry Medical College and Vanderbilt University.
Join the group
I'm accepting master's and Ph.D. students. If you're curious and want to work on real problems with AI, email me or book a meeting.
Publications 39
All of my papers, mirrored daily from my Google Scholar profile. Sort them by date, citations or venue.
Citations
| All | Since 2021 | |
|---|---|---|
| Citations | 207 | 203 |
| h-index | 8 | 8 |
| i10-index | 7 | 7 |
Textbook
Jesse Roberts, 2024, Introduction to PLC Automation, PDF Version
Papers
2026
- Augmenting LLM-Based Agents for Improved Performance in Pentesting and Commissioning Operational Technology in Critical Infrastructure Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States), 2026
- Calibrating Urban Traffic Simulation from Sparse Road Observations via Genetic Optimization arXiv preprint arXiv:2606.03823, 2026
- Exploitation Is All You Need… for Exploration 2026 International Joint Conference on Neural Networks (IJCNN), 1-6, 2026
- Generation and Validation of Configuration Management Code for Cyber Range Environments Using Large Language Models The International FLAIRS Conference Proceedings 39 (1), 2026
- Human-Alignment, Calibration, and Activation Patterns in Large Language Model Uncertainty arXiv preprint arXiv:2605.30675, 2026
- Investigating Human-Aligned Large Language Model Uncertainty The International FLAIRS Conference Proceedings 39 (1), 2026
- KARMA: Karma-Aligned Reward Model Adaptation Findings of the Association for Computational Linguistics: EMNLP 2026, 48030 …, 2026
- Leveraging BART to Assess CS1 C++ Programming Assignments using Rubric-based Criteria arXiv preprint arXiv:2606.03814, 2026
- Multimodal Language Model Vision Analysis for Leader Drone Identification in Adversarial Swarms IEEE Internet of Things Journal, 2026
- Reinforcement learning from meta-evaluation: Aligning language models without ground-truth labels arXiv preprint arXiv:2601.21268, 2026
- Reward-Guided Fine-Tuning of Language Models with Social Feedback The International FLAIRS Conference Proceedings 39 (1), 2026
- SAGE 0.2: LLMs for DOM Informed Internet Guidance The International FLAIRS Conference Proceedings 39 (1), 2026
- Systematic Analysis of Tokenization Properties in Low-Resource Polysynthetic NMT The International FLAIRS Conference Proceedings 39 (1), 2026
- When Prediction Isn’t Enough: Decision-Criteria-Informed Learning for Threshold-Based Decisions 2026 25th International Conference on Machine Learning and Applications …, 2026
2025
- Basic Categories in Vision Language Models: Expert Prompting Doesn’t Grant Expertise Twelfth Annual Conference on Advances in Cognitive Systems 2025, 177, 2025
- Do Large Language Models Learn Human-Like Strategic Preferences? Proceedings of the 1st Workshop for Research on Agent Language Models (REALM …, 2025
- Enhancing Post Earnings Announcement Drift Measurement with Large Language Models Proceedings of The 10th Workshop on Financial Technology and Natural …, 2025
- Harnessing the Power of AI in Honors: Equipping Students to Leverage Tools Effectively Journal of the National Collegiate Honors Council 26 (1), 51-59, 2025
- Leveraging large language models and machine learning for smart contract vulnerability detection 2025 IEEE 15th Annual Computing and Communication Workshop and Conference …, 2025
- LLMs as Agentic Cooperative Players in Multiplayer UNO International Conference on Machine Learning Applications 2025, 2025
- RL+ Transformer= A General-Purpose Problem Solver Proceedings of the 1st Workshop for Research on Agent Language Models (REALM …, 2025
- Tennessee Tech University: Action Plan for Artificial Intelligence Education, Workforce Development, Research, and Infrastructure Needs 2025
- Toward the Use of LLMs to Support Curriculum Mapping to Established Frameworks American Society of Engineering Education, 2025
2024
- A Theoretical & Empirical Analysis of Transformer Language Model Behavior 2024
- Chain of thought still thinks fast: Apricot helps with thinking slow Proceedings of the Annual Meeting of the Cognitive Science Society 47, 2024
- Do Large Language Models Learn to Human-Like Learn? Proceedings of the AAAI 2024 Spring Symposium Series 3 (1), 588-591, 2024
- Investigating Expert-in-the-Loop LLM Discourse Patterns for Ancient Intertextual Analysis 4th International Conference on Natural Language Processing for Digital …, 2024
- Large language model recall uncertainty is modulated by the fan effect 28th Conference on Computational Natural Language Learning, 303–313, 2024
- Supporting the digital autonomy of elders through llm assistance Proceedings of the AAAI Symposium Series 4 (1), 182-186, 2024
- The base-rate effect on llm benchmark performance: Disambiguating test-taking strategies from benchmark performance Findings of the Association for Computational Linguistics: EMNLP 2024, 2283–2288, 2024
- Using artificial populations to study psychological phenomena in neural models Proceedings of the aaai conference on artificial intelligence 38 (17), 18906 …, 2024
2023
- How Powerful are Decoder-Only Transformer Neural Models? 2024 International Joint Conference on Neural Networks (IJCNN), 1--8, 2023
2021
- Finding an Equilibrium in the Traveler's Dilemma with Fuzzy Weak Domination 2021 IEEE Conference on Games (CoG), 1-5, 2021
2020
- Extending the Philosophy of Computational Criticism. ICCC, 358-361, 2020
- pReview: The artificially intelligent conference reviewer 2020 19th IEEE International Conference on Machine Learning and Applications …, 2020
2019
- Biologically Extending the Gen 2 ANN Model. FLAIRS, 132-135, 2019
2017
- Improving any arbitrary MPPT hill climber with ANN estimations 2017 IEEE 44th Photovoltaic Specialist Conference (PVSC), 3083-3087, 2017
- MNFIS+; or, a Better Hybrid Heuristic Maximum Power Point Tracker Tennessee Technological University, 2017
2016
- MNFIS and other soft computing based MPPT techniques: A comparative analysis 2016 IEEE 43rd Photovoltaic Specialists Conference (PVSC), 3247-3251, 2016
Mirrored from Google Scholar, last updated November 8, 2026.