LLMs help clinicians judge equations from symbolic regression
LLMs as Post-hoc Auditors of Physiological Plausibility in Symbolic Regression: A Clinician-Evaluated Case Study
Artificial Intelligence
Summary
Finding simple, understandable formulas from complex health data is very useful but can be tricky because some formulas can be too complicated or not make sense medically. The authors studied if language AI models can help doctors by reviewing and comparing these formulas to see which ones make more sense for predicting body fat. They found that AI models are better at comparing different formulas than explaining each part perfectly. However, doctors still need to check the AI’s work because it can sometimes give wrong or confusing ideas.
What this means in practice
- •For clinical data scientists: Use LLMs to rank and compare mathematical models from symbolic regression for medical data, helping clinicians select more plausible formulas.
- •For healthcare ai developers: Incorporate LLM-based auditing tools in AI systems to provide comparative assessments of model interpretability under expert supervision.
Tested on one dataset.
Authors
Jorge López-Varela, J. Ignacio Hidalgo, José-Manuel Muñoz, Omar Costilla-Reyes, Esther Maqueda, Jesus Moreno-Fernandez, Tomás González-Vidal, J. Manuel Velasco, Oscar Garnica
Abstract
Genetic Programming and its variants, such as grammatical evolution, are widely used in Symbolic Regression to derive mathematical expressions from multivariate data. In addition to predictive accuracy, models are appreciated for their potential to provide interpretability, offering explicit equations that relate input variables to outcomes. However, achieving interpretability and plausibility remains challenging, as evolved models may be complex or scientifically inconsistent. In this study, we explore whether Large Language Models, can assist in improving the explainability of Symbolic Regression models generated by evolutionary computation methods. Building upon our previous work on estimating body fat percentage using grammar-based Genetic Programming , we investigate the use of LLMs as post-processing tools to analyze and rank evolved expressions according to their interpretability and medical plausibility. Four symbolic expressions are analysed by three LLMs over three repeated runs, and the resulting interpretations and rankings are assessed by a panel of three clinicians. Across the three LLMs, comparative model-ranking outputs received more favorable clinician assessments than isolated term-level interpretations. However, the LLMs also produced physiologically and mathematically questionable explanations, indicating that they are better suited to comparative auditing under expert oversight than to autonomous validation.\blfootnote{The present work is an extended version of a paper submitted into a journal.