VArsity: Can Large Language Models Keep Power Engineering Students in Phase?

This paper provides an educational case study regarding our experience in deploying ChatGPT Large Language Models (LLMs) in the Spring 2025 and Fall 2023 offerings of ECE 4320: Power System Analysis & Control at Georgia Tech. As part of course assessments, students were tasked with identifying, explaining, and correcting errors in the ChatGPT outputs corresponding to power factor correction problems. While most students successfully identified the errors in the outputs from the GPT-4 version of ChatGPT used in Fall 2023, students found the errors from the ChatGPT o1 version much more difficult to identify in Spring 2025. As shown in this case study, the role of LLMs in pedagogy, assessment, and learning in power engineering classrooms is an important topic deserving further investigation.

Interpretability cs.CY eess.SY interpretability Engineering Computer Science

Source

Paper

References (17)

01DualSchool: How Reliable are LLMs for Optimization Education?Michael Klamkin, Arnaud Deza, Sikai Cheng et al.2025 · arXiv.org · 5 citations In Library

02Assessing GPT Performance in a Proof-Based University-Level Course Under Blind GradingMing Ding, Rasmus Kyng, Federico Soldà et al.2025 · arXiv.org · 1 citation In Library

03Exploring the Capabilities and Limitations of Large Language Models in the Electric Energy SectorLin Dong, Subir Majumder, Fatemeh Doudi et al.2024 · Joule · 151 citations In Library

04Examining the Potential and Pitfalls of ChatGPT in Science and Engineering Problem-SolvingKaren D. Wang, E. Burkholder, C. Wieman et al.2023 · Frontiers in Education · 81 citations In Library

05ChatGPT for good? On opportunities and challenges of large language models for educationEnkelejda Kasneci, Kathrin Seßler, S. Küchemann et al.2023 · Learning and Individual Differences · 6.3k citations

06Large language models challenge the future of higher educationS. Milano, J. McGrane, S. Leonelli2023 · Nature Machine Intelligence · 164 citations

07A capacitor has a negative reactance, so X c, fixed should be a negative value

08The real and reactive power expressions are not multiplied by the impedance values. Actually, this whole portion of the response is not needed

09The problem does not have any transformers, so tap ratios do not need to be considered, and there are also no shunt elements. Finally

10The switched capacitor’s reactance is based on the difference between the fixed capacitor’s reactance and the desired reactance

11parallel with the load and voltage source) Switched capacitor Not used

12The load impedance is Z = (10 × 10 3 ) 2 / (10 × 10 6 + j 5 × 10 6 ) ∗ = 8+ j 4 Ω ) The reactive power consumed by a shunt capacitance is Q = −| V | 2 / ( X c ) ,

Scroll for more · 5 remaining