Comparative Evaluation of Prompting and Fine-Tuning for Applying Large Language Models to Grid-Structured Geospatial Data

TL;DR

This study compares prompting and fine-tuning methods for large language models in interpreting grid-structured geospatial data, revealing their respective strengths and limitations in spatial-temporal reasoning tasks.

Contribution

It provides a systematic evaluation of prompting versus fine-tuning for LLMs applied to geospatial data, highlighting the advantages of fine-tuning for complex reasoning.

Findings

Fine-tuning improves accuracy in structured geospatial reasoning.

Prompting performs well in zero-shot scenarios but has limitations.

Fine-tuned models excel in temporal and spatial reasoning tasks.

Abstract

This paper presents a comparative study of large language models (LLMs) in interpreting grid-structured geospatial data. We evaluate the performance of a base model through structured prompting and contrast it with a fine-tuned variant trained on a dataset of user-assistant interactions. Our results highlight the strengths and limitations of zero-shot prompting and demonstrate the benefits of fine-tuning for structured geospatial and temporal reasoning.