Language Models for Multimessenger Astronomy

With the increasing reliance of astronomy on multi-instrument and multi-messenger observations for detecting transient phenomena, communication among astronomers has become more critical. Apart from automatic prompt follow-up observations, short reports, e.g., GCN circulars ATels, provide essential human-written interpretations discussions observations. These reports lack a defined format, unlike machine-readable messages, making it challenging to associate phenomena with specific objects or coordinates in sky. This paper examines use large language models (LLMs)—machine learning billions trainable parameters that are trained text—such as InstructGPT-3 open-source Flan-T5-XXL extracting information astronomical reports. The study investigates zero-shot few-shot capabilities LLMs demonstrates various techniques improve accuracy predictions. shows importance careful engineering while working LLMs, demonstrated through edge case examples. study’s findings have significant implications development data-driven applications astrophysical text analysis.