Prototype-based Explainable Neural Networks with Channel-specific Reasoning for Geospatial Learning Tasks Journal: Environmental Data Science Publication type: Methods Paper
Anushka Narayanan, Karianne J. Bergen · 2026
Read Prototype-based Explainable Neural Networks with Channel-specific Reasoning for Geospatial Learning Tasks Journal: Environmental Data Science Publication type: Methods Paper by Anushka Narayanan, Karianne J. Bergen (2026) 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-2602-00331.txt.
Description
Anushka Narayanan and Karianne J. Bergen develop prototype based explainable networks for geospatial tasks. Such AI matters for understanding machine decisions and ensuring model trust in science. These methods predict from similarity between inputs and learned prototypes for typical classes, offering clear readings over shaky post hoc explanations. Existing models target standard RGB images, not the distinct channels of earth science image and raster data. Their approach tailors this reasoning to multi channel geo data, where each channel stands for a distinct physical variable or spectral channel.
Catalog record
| CREATOR | Anushka Narayanan, Karianne J. Bergen |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 78679 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602.00331