STEM Department

Director: Cade Konishi

Email: cadekonishi@gmail.com

Research on STEM engagement, curriculum design, and youth participation in science and technology.

Artificial Intelligence in STEM Education

Education initiatives, curriculum design, and community programs. Click a paper to view full images and supplementary figures.

Constraint Drift in Large Language Models: Evaluating Reliability in Engineering Reasoning

Eileen Shao - 2026

Large language models (LLMs) are increasingly being used in engineering related contexts such as education, coding environments, and design assistance tools. While they are effective at generating explanations, solving structured problems, and supporting early stage ideation, their reliability in engineering reasoning tasks that involve multiple interacting constraints remains uncertain. Unlike general benchmark questions, engineering problems require constraints to remain consistent across multi-step reasoning processes rather than being applied only at the final step. Recent research in both general reasoning benchmarks and STEM-focused datasets suggests that performance decreases as task complexity and constraint interactions increase. This paper synthesizes findings from existing literature on both general LLM reasoning and STEM-specific evaluation benchmarks. Across these studies, a consistent pattern emerges: while models perform well on individual reasoning steps, their consistency declines when constraints must be maintained over longer reasoning chains. To describe this behavior, the concept of constraint drift is introduced, referring to the gradual loss of constraint consistency during multi-step reasoning. This framework is used to better understand limitations in current evaluation methods for engineering processes and applications.

Materials Informatics: Accelerating Materials Discovery Through Artificial Intelligence

Cade Konishi - 2026

Materials informatics is the application of artificial intelligence, data science, and machine learning to materials science. In contrast to traditional materials discovery, which relies on trial-and-error, materials informatics uses existing datasets of material properties to predict the behavior of new materials before they are synthesized. By applying AI to materials science, materials discovery becomes much faster and more efficient. Materials informatics is used in a large number of industries, and applications include but are not limited to battery material discovery, polymer design, and semiconductors design. Though materials informatics is largely beneficial for materials discovery, small datasets, model interpretability, and conflicting real-world requirements remain challenges in the field. Overall, materials informatics is a powerful tool for accelerating materials discovery and development.