ChatGPT-4o, when given lunar sample report text, extracts oxide and element weight ranges whose midpoints mostly agree with manual ground truth within 5% relative error, and it outperforms the same model queried without the report.
A survey on RAG meeting LLMs: Towards retrieval-augmented large language models
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.DL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Towards Large Language Models for Lunar Mission Planning and In Situ Resource Utilization
ChatGPT-4o, when given lunar sample report text, extracts oxide and element weight ranges whose midpoints mostly agree with manual ground truth within 5% relative error, and it outperforms the same model queried without the report.