Saved in:
Bibliographic Details
Main Authors: Singh, Anshul, Chaudhary, Rohan, Singh, Gagneet, Kumary, Abhay
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2511.17238
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915630918139904
author Singh, Anshul
Chaudhary, Rohan
Singh, Gagneet
Kumary, Abhay
author_facet Singh, Anshul
Chaudhary, Rohan
Singh, Gagneet
Kumary, Abhay
contents The impressive performance of VLMs is largely measured on benchmarks that fail to capture the complexities of real-world scenarios. Existing datasets for tabular QA, such as WikiTableQuestions and FinQA, are overwhelmingly monolingual (English) and present tables in a digitally perfect, clean format. This creates a significant gap between research and practice. To address this, we present \textbf{MirageTVQA}, a new benchmark designed to evaluate VLMs on these exact dimensions. Featuring nearly 60,000 QA pairs across 24 languages, MirageTVQA challenges models with tables that are not only multilingual but also visually imperfect, incorporating realistic noise to mimic scanned documents. Our evaluation of the leading VLMs reveals two primary failure points: a severe degradation in performance (over 35\% drop for the best models) when faced with visual noise and a consistent English-first bias where reasoning abilities fail to transfer to other languages. MirageTVQA provides a benchmark for measuring and driving progress towards more robust VLM models for table reasoning. The dataset and the code are available at: https://github.com/anshulsc/MirageTVQA.
format Preprint
id arxiv_https___arxiv_org_abs_2511_17238
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Lost in Translation and Noise: A Deep Dive into the Failure Modes of VLMs on Real-World Tables
Singh, Anshul
Chaudhary, Rohan
Singh, Gagneet
Kumary, Abhay
Computation and Language
Artificial Intelligence
Computer Vision and Pattern Recognition
The impressive performance of VLMs is largely measured on benchmarks that fail to capture the complexities of real-world scenarios. Existing datasets for tabular QA, such as WikiTableQuestions and FinQA, are overwhelmingly monolingual (English) and present tables in a digitally perfect, clean format. This creates a significant gap between research and practice. To address this, we present \textbf{MirageTVQA}, a new benchmark designed to evaluate VLMs on these exact dimensions. Featuring nearly 60,000 QA pairs across 24 languages, MirageTVQA challenges models with tables that are not only multilingual but also visually imperfect, incorporating realistic noise to mimic scanned documents. Our evaluation of the leading VLMs reveals two primary failure points: a severe degradation in performance (over 35\% drop for the best models) when faced with visual noise and a consistent English-first bias where reasoning abilities fail to transfer to other languages. MirageTVQA provides a benchmark for measuring and driving progress towards more robust VLM models for table reasoning. The dataset and the code are available at: https://github.com/anshulsc/MirageTVQA.
title Lost in Translation and Noise: A Deep Dive into the Failure Modes of VLMs on Real-World Tables
topic Computation and Language
Artificial Intelligence
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2511.17238