TerraMARS: A Domain-Adapted Small-Language-Model Pipeline for Mars Terraforming Literature

2026-06-17

AI Summary

This study addresses the challenge posed by the unstructured nature of Mars science literature, which hinders its effective use in habitability assessments and terraforming research. The work proposes the first end-to-end information extraction framework tailored to this domain, leveraging the Google Gemma 3 1B model enhanced with QLoRA fine-tuning, multi-stage retrieval, and text chunking strategies to automatically convert open-access Mars science papers into structured JSON knowledge. Notably, this is the first application of quantized low-rank adaptation (QLoRA) to planetary science texts. The resulting structured knowledge base supports downstream tasks such as question answering and digital twin modeling, establishing a computable foundation for Mars research. Nevertheless, the authors acknowledge that factual consistency and extraction accuracy require further improvement.

Abstract

Researchers are interested in learning about Mars so that it may eventually become habitable for humans. To achieve this, there is a need for comprehensive knowledge of the planet's atmosphere, hydrology, surface chemistry, radiation environment, and spatial features through the scientific literature. These contain valuable information and meaningful quantitative constraints that can be used in other models and studies, such as habitability assessment and future terraforming studies. We present TerraMARS, an end-to-end information extraction pipeline that combines a domain-adapted Small Language Model to answer Mars terraforming-related questions and convert unstructured Mars science text into machine-readable structured outputs in JavaScript Object Notation (JSON) format. A corpus of open-access papers is collected and processed using a multistage retrieval and chunking framework. Google Gemma 3 1B was adapted to the domain using Quantized Low-Rank Adaptation (QLoRA) fine-tuning on Mars-specific question-answering and information extraction datasets. The resulting pipeline generates both types of output and provides a foundation for integrating knowledge from scientific literature into downstream applications like digital twins and habitability modeling for Mars. The output from this pipeline looks promising, but further improvements are needed to increase extraction accuracy and factual consistency.

Problem

• Mars terraforming
• scientific literature
• information extraction
• habitable environment
• structured data

Innovation

• Small Language Model
• Domain Adaptation
• QLoRA
• Information Extraction
• Mars Terraforming