Saved in:
Bibliographic Details
Main Authors: Seo, Jean, Lee, Hyunkyung, Kim, Gibaeg, Han, Wooseok, Yoo, Jaehyo, Lim, Seungseop, Shin, Kihun, Yang, Eunho
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2509.22041
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908570674528256
author Seo, Jean
Lee, Hyunkyung
Kim, Gibaeg
Han, Wooseok
Yoo, Jaehyo
Lim, Seungseop
Shin, Kihun
Yang, Eunho
author_facet Seo, Jean
Lee, Hyunkyung
Kim, Gibaeg
Han, Wooseok
Yoo, Jaehyo
Lim, Seungseop
Shin, Kihun
Yang, Eunho
contents Safety is a paramount concern in clinical chatbot applications, where inaccurate or harmful responses can lead to serious consequences. Existing methods--such as guardrails and tool calling--often fall short in addressing the nuanced demands of the clinical domain. In this paper, we introduce TACOS (TAxonomy of COmprehensive Safety for Clinical Agents), a fine-grained, 21-class taxonomy that integrates safety filtering and tool selection into a single user intent classification step. TACOS is a taxonomy that can cover a wide spectrum of clinical and non-clinical queries, explicitly modeling varying safety thresholds and external tool dependencies. To validate our taxonomy, we curate a TACOS-annotated dataset and perform extensive experiments. Our results demonstrate the value of a new taxonomy specialized for clinical agent settings, and reveal useful insights about train data distribution and pretrained knowledge of base models.
format Preprint
id arxiv_https___arxiv_org_abs_2509_22041
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Taxonomy of Comprehensive Safety for Clinical Agents
Seo, Jean
Lee, Hyunkyung
Kim, Gibaeg
Han, Wooseok
Yoo, Jaehyo
Lim, Seungseop
Shin, Kihun
Yang, Eunho
Computation and Language
Safety is a paramount concern in clinical chatbot applications, where inaccurate or harmful responses can lead to serious consequences. Existing methods--such as guardrails and tool calling--often fall short in addressing the nuanced demands of the clinical domain. In this paper, we introduce TACOS (TAxonomy of COmprehensive Safety for Clinical Agents), a fine-grained, 21-class taxonomy that integrates safety filtering and tool selection into a single user intent classification step. TACOS is a taxonomy that can cover a wide spectrum of clinical and non-clinical queries, explicitly modeling varying safety thresholds and external tool dependencies. To validate our taxonomy, we curate a TACOS-annotated dataset and perform extensive experiments. Our results demonstrate the value of a new taxonomy specialized for clinical agent settings, and reveal useful insights about train data distribution and pretrained knowledge of base models.
title Taxonomy of Comprehensive Safety for Clinical Agents
topic Computation and Language
url https://arxiv.org/abs/2509.22041