# From Canonical to Complex: Benchmarking LLM Capabilities in Undergraduate Thermodynamics

## TL;DR

This paper evaluates the ability of large language models to understand and reason about undergraduate thermodynamics concepts, revealing current limitations in their competence for unsupervised educational use.

## Contribution

The authors introduce UTQA, a comprehensive thermodynamics question benchmark, and analyze LLM performance, highlighting gaps in reasoning and visual understanding.

## Key findings

- No model exceeded 95% accuracy threshold.
- Best models achieved 82% accuracy, mostly on text-only questions.
- Performance drops significantly on visual reasoning and irreversible processes.

## Abstract

Large language models (LLMs) are increasingly considered as tutoring aids in science education. Yet their readiness for unsupervised use in undergraduate instruction remains uncertain, as reliable teaching requires more than fluent recall: it demands consistent, principle-grounded reasoning. Thermodynamics, with its compact laws and subtle distinctions between state and path functions, reversibility, and entropy, provides an ideal testbed for evaluating such capabilities. Here we present UTQA, a 50-item undergraduate thermodynamics question answering benchmark, covering ideal-gas processes, reversibility, and diagram interpretation. No leading 2025-era model exceeded our 95\% competence threshold: the best LLMs achieved 82\% accuracy, with text-only items performing better than image reasoning tasks, which often fell to chance levels. Prompt phrasing and syntactic complexity showed modest to little correlation with performance. The gap concentrates in finite-rate/irreversible scenarios and in binding visual features to thermodynamic meaning, indicating that current LLMs are not yet suitable for unsupervised tutoring in this domain.