Table Understanding and (Multimodal) LLMs: A Cross-Domain Case Study on Scientific vs. Non-Scientific Data
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866911121353474048 |
|---|---|
| author | Borisova, Ekaterina Barth, Fabio Feldhus, Nils Ahmad, Raia Abu Ostendorff, Malte Suarez, Pedro Ortiz Rehm, Georg Möller, Sebastian |
| author_facet | Borisova, Ekaterina Barth, Fabio Feldhus, Nils Ahmad, Raia Abu Ostendorff, Malte Suarez, Pedro Ortiz Rehm, Georg Möller, Sebastian |
| contents | Tables are among the most widely used tools for representing structured data in research, business, medicine, and education. Although LLMs demonstrate strong performance in downstream tasks, their efficiency in processing tabular data remains underexplored. In this paper, we investigate the effectiveness of both text-based and multimodal LLMs on table understanding tasks through a cross-domain and cross-modality evaluation. Specifically, we compare their performance on tables from scientific vs. non-scientific contexts and examine their robustness on tables represented as images vs. text. Additionally, we conduct an interpretability analysis to measure context usage and input relevance. We also introduce the TableEval benchmark, comprising 3017 tables from scholarly publications, Wikipedia, and financial reports, where each table is provided in five different formats: Image, Dictionary, HTML, XML, and LaTeX. Our findings indicate that while LLMs maintain robustness across table modalities, they face significant challenges when processing scientific tables. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_00152 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Table Understanding and (Multimodal) LLMs: A Cross-Domain Case Study on Scientific vs. Non-Scientific Data Borisova, Ekaterina Barth, Fabio Feldhus, Nils Ahmad, Raia Abu Ostendorff, Malte Suarez, Pedro Ortiz Rehm, Georg Möller, Sebastian Computation and Language Tables are among the most widely used tools for representing structured data in research, business, medicine, and education. Although LLMs demonstrate strong performance in downstream tasks, their efficiency in processing tabular data remains underexplored. In this paper, we investigate the effectiveness of both text-based and multimodal LLMs on table understanding tasks through a cross-domain and cross-modality evaluation. Specifically, we compare their performance on tables from scientific vs. non-scientific contexts and examine their robustness on tables represented as images vs. text. Additionally, we conduct an interpretability analysis to measure context usage and input relevance. We also introduce the TableEval benchmark, comprising 3017 tables from scholarly publications, Wikipedia, and financial reports, where each table is provided in five different formats: Image, Dictionary, HTML, XML, and LaTeX. Our findings indicate that while LLMs maintain robustness across table modalities, they face significant challenges when processing scientific tables. |
| title | Table Understanding and (Multimodal) LLMs: A Cross-Domain Case Study on Scientific vs. Non-Scientific Data |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2507.00152 |