WCN26-4700 Robustness Gap of Large Language Models in Nephrology

Read full text Summarise with VIRAT
Authors: Ayaka Soejima, Fumiya Kitano, Mamoru Masaki, Daisuke Ichikawa, Yugo Shibagaki, Ryunosuke Noda
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Large language models (LLMs) achieve high accuracy on medical benchmarks, raising interest in their clinical application. However, whether this performance reflects genuine reasoning or pattern recognition remains unclear. To evaluate reasoning robustness, we replaced the correct answer in nephrology multiple-choice questions with “None of the other answers” (NOTA) and assessed changes in accuracy. We hypothesized that causal and pathophysiological reasoning would preserve accuracy, whereas reliance on memorized patterns would cause a marked decline.

Fields of Study & Subjects

Artificial Intelligence in Healthcare and Education Machine Learning in Healthcare Clinical Reasoning and Diagnostic Skills

Publication details

Journal / source
Kidney International Reports
Volume
11
Issue
4
Pages
105647–105647
Publication date
25 March 2026
Language
EN