A list of my research papers can be seen on my Google Scholar page.
IEEE Working Group P3109 Interim Report on Arithmetic Formats for Machine Learning
K. Gunnam, L. Tsai, J. Sarnoff, A. Fitzgibbon, C. Wintersteiger, A. Omondi, G. Lemieux, A. Mushi, A. Naseer, A. Martin, A. Zakharchenko, A. Sazegari, B. Noune, C. Luschi, D. Huang, D. Chen, D. Lutz, E. Schwarz, G. Tabak, J. Davenport, J. Demmel, J. Riedy, J. Gustafson, J. Villalba, K. Parry, K. Docker, K. Wong, L. Hunhold, M. Mikaitis, M. Cococcioni, M. Cornea, M. Fasi, M. Overton, M. Siu, M. Hack, M. Cowlishaw, N. Revol, N. Badizadegan, O. Sentieys, P. Balanca, S. Nagarakatte, S. Ko, S. Filip, S. Oberman, T. Yeh, T. Ly. IEEE Working Group P3109 Interim Report on Binary Floating-point Formats for Machine Learning, Sep 2026.
Be Like Water: Adaptive Floating Point for Machine Learning
Thomas Yeh, Max Sterner, Zerlina Lai, Brandon Chuang and Alexander Ihler
Proceedings of the 39th International Conference on Machine Learning, 2022.
Our study presents a tool to help students reclaim productive struggle.
Learning Sentinel: A Metacognition-Driven Tool for Self-Regulated AI Use
Jared Apillanes, Candice Lu, Mason Lam, Jin Woo Choi, Seth Flucas, Xingyu Li, Jennifer Wong-Ma, Thomas Y. Yeh
In Proceedings of the 58th ACM Technical Symposium on Computer Science Education (SIGCSE TS 2027).
Our study presents a new taxonomy to evaluate cognitive offloading of student-AI interactions.
Our paper won the Best Paper Award at SIGCSE Virtual 2026!
Chat-Bloom: A Taxonomy for Classifying Cognitive Offloading in Student-AI interactions in CS Education
Leo Chao, Christiana Cho, Candice Lu, Mikayla McGrath, Troy Yeh, Misaki Rios, Ellie Yeh, Sowmiyalakshmi Ganesh and Thomas Y. Yeh
In Proceedings of the 2nd ACM Virtual Global Computing Education Conference (SIGCSE Virtual 2026).
Our study presents statistical evidence on students' immediate learning and knowledge retention when using commercial vs pedagogical AI tutors.
Pacing for Mastery: Optimizing LLM Interactions for Learning
Karena Tran, Ge Gao, Angela Lombard, Tyler Yu, Haoning Jiang and Thomas Yeh
In Proceedings of the 57th ACM Technical Symposium on Computer Science
Education V. 1 (SIGCSE TS 2026).
Our study compares AI-assisted to manual grading of handwritten CS assessments.
Fighting Fire with Fire: LLM-Assisted Grading of Handwritten CS Assessments
Jared Apillanes, Jason Weber, Sergio Gago-Masague, Jennifer Wong-Ma and Thomas Yeh
In Proceedings of the 57th ACM Technical Symposium on Computer Science
Education V. 1 (SIGCSE TS 2026).
As LLMs become integral tools for experienced programmers, their impact on novice learners remains a critical question. Our study delves into this challenge, revealing how interactive LLMs can enhance code generation accuracy for beginners and improve their prompting skills. Our approach not only boosts learning outcomes but also addresses equity concerns in CS education.
Bridging Novice Programmers and LLMs with Interactivity
Thomas Y. Yeh, Karena Tran, Ge Gao, Tyler Yu, Wai On Fong, and Tzu-Yi Chen.
2025. Bridging Novice Programmers and LLMs with Interactivity.
In Proceedings of the 56th ACM Technical Symposium on Computer Science
Education V. 1 (SIGCSE TS 2025). ACM, New York, NY, USA, 7 pages.
https://doi.org/10.1145/3641554.3701867