LLMs Evaluate Physiological Plausibility in Symbolic Regression
September 11, 2026
This study tests using LLMs as post-hoc auditors to rank the interpretability and medical plausibility of mathematical expressions generated via Genetic Programming. The approach helps bridge the gap between predictive accuracy and scientific consistency in symbolic regression models.
HOW THIS AFFECTS YOU
●
researcherYou can leverage LLMs to automate the qualitative validation of evolved mathematical models.
●
healthThis provides a framework for ensuring that AI-derived medical equations are biologically and clinically sound.