Saved in:
Bibliographic Details
Main Authors: Toles, Matthew, Singh, Rattandeep, Song, Isaac, Yu, Zhou
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2506.14079
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914271634391040
author Toles, Matthew
Singh, Rattandeep
Song, Isaac
Yu, Zhou
author_facet Toles, Matthew
Singh, Rattandeep
Song, Isaac
Yu, Zhou
contents Completing paperwork is a challenging and time-consuming problem. Form filling is especially challenging in the pure-image domain without access to OCR, typeset PDF text, or a DOM. For computer agents, it requires multiple abilities, including multi-modal understanding, information retrieval, and tool-use. We present a novel form-filling benchmark consisting of 432 fields spread across 55 documents and 3 tasks, requiring knowledge of 236 features per user. We find that baseline VLAs achieve less than 1% accuracy in most cases, primarily due to poor localization ability. GUI agents also struggle, scoring between 10.6-68.0% despite high cost and latency. Therefore, we also contribute FieldFinder, a tool to assist LLMs in identifying where to place text on a form. With FieldFinder, all models achieve equal or better performance in all six study conditions, with a maximum increase from 2% to 56%.
format Preprint
id arxiv_https___arxiv_org_abs_2506_14079
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle FormGym: Doing Paperwork with Agents
Toles, Matthew
Singh, Rattandeep
Song, Isaac
Yu, Zhou
Artificial Intelligence
Completing paperwork is a challenging and time-consuming problem. Form filling is especially challenging in the pure-image domain without access to OCR, typeset PDF text, or a DOM. For computer agents, it requires multiple abilities, including multi-modal understanding, information retrieval, and tool-use. We present a novel form-filling benchmark consisting of 432 fields spread across 55 documents and 3 tasks, requiring knowledge of 236 features per user. We find that baseline VLAs achieve less than 1% accuracy in most cases, primarily due to poor localization ability. GUI agents also struggle, scoring between 10.6-68.0% despite high cost and latency. Therefore, we also contribute FieldFinder, a tool to assist LLMs in identifying where to place text on a form. With FieldFinder, all models achieve equal or better performance in all six study conditions, with a maximum increase from 2% to 56%.
title FormGym: Doing Paperwork with Agents
topic Artificial Intelligence
url https://arxiv.org/abs/2506.14079