Enabling Intrinsic Reasoning over Dense Geospatial Embeddings with DFR-Gemma

TL;DR

This paper introduces DFR-Gemma, a framework enabling LLMs to directly reason over dense geospatial embeddings, improving efficiency and accuracy in geospatial intelligence tasks.

Contribution

It presents a novel method to align high-dimensional geospatial embeddings with LLMs, eliminating the need for textual conversion and enhancing reasoning capabilities.

Findings

DFR-Gemma enables zero-shot reasoning over spatial features.

The approach significantly improves efficiency over text-based methods.

Experimental results show accurate decoding of spatial patterns.

Abstract

Representation learning for geospatial and spatio-temporal data plays a critical role in enabling general-purpose geospatial intelligence. Recent geospatial foundation models, such as the Population Dynamics Foundation Model (PDFM), encode complex population and mobility dynamics into compact embeddings. However, their integration with Large Language Models (LLMs) remains limited. Existing approaches to LLM integration treat these embeddings as retrieval indices or convert them into textual descriptions for reasoning, introducing redundancy, token inefficiency, and numerical inaccuracies. We propose Direct Feature Reasoning-Gemma (DFR-Gemma), a novel framework that enables LLMs to reason directly over dense geospatial embeddings. DFR aligns high-dimensional embeddings with the latent space of an LLM via a lightweight projector, allowing embeddings to be injected as semantic tokens alongside natural language instructions. This design eliminates the need for intermediate textual representations and enables intrinsic reasoning over spatial features. To evaluate this paradigm, we introduce a multi-task geospatial benchmark that pairs embeddings with diverse question-answer tasks, including feature querying, comparison, and semantic description. Experimental results show that DFR allows LLMs to decode latent spatial patterns and perform accurate zero-shot reasoning across tasks, while significantly improving efficiency compared to text-based baselines. Our results demonstrate that treating embeddings as primary data inputs, provides a more direct, efficient, and scalable approach to multimodal geospatial intelligence.