Assessing Y-Axis Influence: Bias in Multimodal Language Models on Chart-to-Table Translation

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Song, Seok Hwan, Efat, Azher Ahmed, Tavanapong, Wallapak
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866918470747160576
author Song, Seok Hwan
Efat, Azher Ahmed
Tavanapong, Wallapak
author_facet Song, Seok Hwan
Efat, Azher Ahmed
Tavanapong, Wallapak
contents Chart-to-table translation converts chart images into structured tabular data. Accurate translation is crucial for Multimodal Language Model (MLM) to answer complex queries. We observe imbalances in the number of images across different aspects of the y-axis information in public chart datasets. Such imbalances can introduce unintended biases, causing uneven MLM performance. Previous works have not systematically examined these biases. To address this gap, we propose a new framework, FairChart2Table, for analyzing y-axis-related bias on five state-of-the-art models. Key Findings: (1) There are significant y-axis biases related to the digit length of the major tick values, the number of major ticks, the range of values, and the tick value format (e.g., abbreviation or scientific format). (2) The number of legends/entities in chart images impacts MLM performance. (3) Prompting MLM with y-axis information can significantly enhance the performance for some MLMs.
format Preprint
id arxiv_https___arxiv_org_abs_2604_24987
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Assessing Y-Axis Influence: Bias in Multimodal Language Models on Chart-to-Table Translation
Song, Seok Hwan
Efat, Azher Ahmed
Tavanapong, Wallapak
Artificial Intelligence
Chart-to-table translation converts chart images into structured tabular data. Accurate translation is crucial for Multimodal Language Model (MLM) to answer complex queries. We observe imbalances in the number of images across different aspects of the y-axis information in public chart datasets. Such imbalances can introduce unintended biases, causing uneven MLM performance. Previous works have not systematically examined these biases. To address this gap, we propose a new framework, FairChart2Table, for analyzing y-axis-related bias on five state-of-the-art models. Key Findings: (1) There are significant y-axis biases related to the digit length of the major tick values, the number of major ticks, the range of values, and the tick value format (e.g., abbreviation or scientific format). (2) The number of legends/entities in chart images impacts MLM performance. (3) Prompting MLM with y-axis information can significantly enhance the performance for some MLMs.
title Assessing Y-Axis Influence: Bias in Multimodal Language Models on Chart-to-Table Translation
topic Artificial Intelligence
url https://arxiv.org/abs/2604.24987