@inproceedings{zhan-etal-2026-reimbursebench,
title = "{R}eimburse{B}ench: Benchmarking {LMM}s as Full-Workflow Auditors for Financial Internal Control",
author = "Zhan, Libin and
Mao, Wuyou and
Li, Xiang and
Zhang, Enci and
Yang, Qi and
Zhu, Shujian and
Hu, Yingtian and
Ye, Wentao and
Wang, Haobo and
Li, Ziman and
Zhang, Jie and
Dong, Wang and
Chen, Jun",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.821/",
doi = "10.18653/2026.findings-emnlp.821",
pages = "17403--17425",
ISBN = "979-8-89176-462-0",
abstract = "Expense reimbursement internal control is a high-stakes corporate finance task that combines multimodal, visually rich evidence (e.g., receipts and approval forms), firm-specific policy text, and multi-step compliance reasoning. No existing benchmark evaluates large multimodal models (LMMs) in this domain. We introduce ReimburseBench, the first benchmark targeting full-workflow multimodal audit reasoning in corporate internal control. It comprises 151 expert-constructed reimbursement workflows, each packaging an application, receipts and supporting documents, and a firm-specific policy, evaluated by a process-level rubric grounded in twelve auditing logic points. Benchmarking nine frontier and locally deployable systems, we find the strongest attains only 72.5{\%}, and no model exceeds 47.7{\%} on hard workflows. Beyond this aggregate score, risk-aware, process-level evaluation surfaces three structural findings: (i) a persistent gap between binary admissibility judgment and precise quantitative computation; (ii) a systematic lenient bias in monetary predictions that poses deployment risk; and (iii) a performance pattern consistent with compositional-generalization difficulties under realistic workflow load. We provide ReimburseBench as a testbed for advancing LMMs toward trustworthy financial systems. Data and code: https://github.com/ikerzhan/ReimburseBench."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhan-etal-2026-reimbursebench">
<titleInfo>
<title>ReimburseBench: Benchmarking LMMs as Full-Workflow Auditors for Financial Internal Control</title>
</titleInfo>
<name type="personal">
<namePart type="given">Libin</namePart>
<namePart type="family">Zhan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wuyou</namePart>
<namePart type="family">Mao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiang</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Enci</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Qi</namePart>
<namePart type="family">Yang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shujian</namePart>
<namePart type="family">Zhu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yingtian</namePart>
<namePart type="family">Hu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wentao</namePart>
<namePart type="family">Ye</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Haobo</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ziman</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jie</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wang</namePart>
<namePart type="family">Dong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jun</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>Expense reimbursement internal control is a high-stakes corporate finance task that combines multimodal, visually rich evidence (e.g., receipts and approval forms), firm-specific policy text, and multi-step compliance reasoning. No existing benchmark evaluates large multimodal models (LMMs) in this domain. We introduce ReimburseBench, the first benchmark targeting full-workflow multimodal audit reasoning in corporate internal control. It comprises 151 expert-constructed reimbursement workflows, each packaging an application, receipts and supporting documents, and a firm-specific policy, evaluated by a process-level rubric grounded in twelve auditing logic points. Benchmarking nine frontier and locally deployable systems, we find the strongest attains only 72.5%, and no model exceeds 47.7% on hard workflows. Beyond this aggregate score, risk-aware, process-level evaluation surfaces three structural findings: (i) a persistent gap between binary admissibility judgment and precise quantitative computation; (ii) a systematic lenient bias in monetary predictions that poses deployment risk; and (iii) a performance pattern consistent with compositional-generalization difficulties under realistic workflow load. We provide ReimburseBench as a testbed for advancing LMMs toward trustworthy financial systems. Data and code: https://github.com/ikerzhan/ReimburseBench.</abstract>
<identifier type="citekey">zhan-etal-2026-reimbursebench</identifier>
<identifier type="doi">10.18653/2026.findings-emnlp.821</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.821/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>17403</start>
<end>17425</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T ReimburseBench: Benchmarking LMMs as Full-Workflow Auditors for Financial Internal Control
%A Zhan, Libin
%A Mao, Wuyou
%A Li, Xiang
%A Zhang, Enci
%A Yang, Qi
%A Zhu, Shujian
%A Hu, Yingtian
%A Ye, Wentao
%A Wang, Haobo
%A Li, Ziman
%A Zhang, Jie
%A Dong, Wang
%A Chen, Jun
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F zhan-etal-2026-reimbursebench
%X Expense reimbursement internal control is a high-stakes corporate finance task that combines multimodal, visually rich evidence (e.g., receipts and approval forms), firm-specific policy text, and multi-step compliance reasoning. No existing benchmark evaluates large multimodal models (LMMs) in this domain. We introduce ReimburseBench, the first benchmark targeting full-workflow multimodal audit reasoning in corporate internal control. It comprises 151 expert-constructed reimbursement workflows, each packaging an application, receipts and supporting documents, and a firm-specific policy, evaluated by a process-level rubric grounded in twelve auditing logic points. Benchmarking nine frontier and locally deployable systems, we find the strongest attains only 72.5%, and no model exceeds 47.7% on hard workflows. Beyond this aggregate score, risk-aware, process-level evaluation surfaces three structural findings: (i) a persistent gap between binary admissibility judgment and precise quantitative computation; (ii) a systematic lenient bias in monetary predictions that poses deployment risk; and (iii) a performance pattern consistent with compositional-generalization difficulties under realistic workflow load. We provide ReimburseBench as a testbed for advancing LMMs toward trustworthy financial systems. Data and code: https://github.com/ikerzhan/ReimburseBench.
%R 10.18653/2026.findings-emnlp.821
%U https://aclanthology.org/2026.findings-emnlp.821/
%U https://doi.org/10.18653/2026.findings-emnlp.821
%P 17403-17425
Markdown (Informal)
[ReimburseBench: Benchmarking LMMs as Full-Workflow Auditors for Financial Internal Control](https://aclanthology.org/2026.findings-emnlp.821/) (Zhan et al., Findings 2026)
ACL
- Libin Zhan, Wuyou Mao, Xiang Li, Enci Zhang, Qi Yang, Shujian Zhu, Yingtian Hu, Wentao Ye, Haobo Wang, Ziman Li, Jie Zhang, Wang Dong, and Jun Chen. 2026. ReimburseBench: Benchmarking LMMs as Full-Workflow Auditors for Financial Internal Control. In Findings of the Association for Computational Linguistics: EMNLP 2026, pages 17403–17425, Budapest, Hungary. Association for Computational Linguistics.