Spatial ModernBERT: Spatial-Aware Transformer for Table and Key-Value Extraction in Financial Documents at Scale
Amrendra Singh [email protected], Maulik Shah [email protected], Dharshan Sampath * · 2025
Read Spatial ModernBERT: Spatial-Aware Transformer for Table and Key-Value Extraction in Financial Documents at Scale by Amrendra Singh [email protected], Maulik Shah [email protected], Dharshan Sampath * (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-08865.txt.
Description
A space-aware reader for finance forms is built by Amrendra Singh and co-authors in a 2025 arXiv paper. Tables and key-value pairs hide in messy bills and reports. Their model adds place codes to a text net so word spots count, not just word sense. Three heads sort each token: field name, column spot, and row start.
Preps on a big table set came first, then fine-tuning on finance forms. A merge step rebuilds the table grid and lifts out the pairs. Text plus place cues beat text alone in their tests. Audit and bill work stand to gain.
Catalog record
| CREATOR | Amrendra Singh [email protected], Maulik Shah [email protected], Dharshan Sampath * |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 39 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-08865