Saved in:
Bibliographic Details
Main Authors: Seo, Jean, Kim, Gibaeg, Shin, Kihun, Lim, Seungseop, Lee, Hyunkyung, Han, Wooseok, Lee, Jongwon, Yang, Eunho
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2601.03627
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917482547118080
author Seo, Jean
Kim, Gibaeg
Shin, Kihun
Lim, Seungseop
Lee, Hyunkyung
Han, Wooseok
Lee, Jongwon
Yang, Eunho
author_facet Seo, Jean
Kim, Gibaeg
Shin, Kihun
Lim, Seungseop
Lee, Hyunkyung
Han, Wooseok
Lee, Jongwon
Yang, Eunho
contents We introduce EPAG, a benchmark dataset and framework designed for Evaluating the Pre-consultation Ability of LLMs using diagnostic Guidelines. LLMs are evaluated directly through HPI-diagnostic guideline comparison and indirectly through disease diagnosis. In our experiments, we observe that small open-source models fine-tuned with a well-curated, task-specific dataset can outperform frontier LLMs in pre-consultation. Additionally, we find that increased amount of HPI (History of Present Illness) does not necessarily lead to improved diagnostic performance. Further experiments reveal that the language of pre-consultation influences the characteristics of the dialogue. By open-sourcing our dataset and evaluation pipeline on https://github.com/seemdog/EPAG, we aim to contribute to the evaluation and further development of LLM applications in real-world clinical settings.
format Preprint
id arxiv_https___arxiv_org_abs_2601_03627
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Evaluating the Pre-Consultation Ability of LLMs using Diagnostic Guidelines
Seo, Jean
Kim, Gibaeg
Shin, Kihun
Lim, Seungseop
Lee, Hyunkyung
Han, Wooseok
Lee, Jongwon
Yang, Eunho
Computation and Language
Artificial Intelligence
We introduce EPAG, a benchmark dataset and framework designed for Evaluating the Pre-consultation Ability of LLMs using diagnostic Guidelines. LLMs are evaluated directly through HPI-diagnostic guideline comparison and indirectly through disease diagnosis. In our experiments, we observe that small open-source models fine-tuned with a well-curated, task-specific dataset can outperform frontier LLMs in pre-consultation. Additionally, we find that increased amount of HPI (History of Present Illness) does not necessarily lead to improved diagnostic performance. Further experiments reveal that the language of pre-consultation influences the characteristics of the dialogue. By open-sourcing our dataset and evaluation pipeline on https://github.com/seemdog/EPAG, we aim to contribute to the evaluation and further development of LLM applications in real-world clinical settings.
title Evaluating the Pre-Consultation Ability of LLMs using Diagnostic Guidelines
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2601.03627