SEMINÁRIO DO DEPARTAMENTO DE ASTRONOMIA
Commercial and Open-Source Large Language Models in Astronomy: Tools, Applications, and Use Cases
a talk by Evgeny Smirnov (Belgrade Observatory, Serbia)
Abstract:
In this talk, I explore how open-source large language models (LLMs) like LLaMA 3.2, Gemma, and DeepSeek can be useful in astronomical data analysis by providing cost-effective alternatives to expensive commercial models such as gpt-4.1/o3 and Claude Sonnet/Opus 4. My research demonstrates that these freely available models, which can run on a researcher's laptop, can achieve acceptable performance comparable to traditional neural networks for astronomical classification tasks — particularly in identifying mean-motion resonances in asteroid dynamics. Through careful prompt engineering, model instruction, and fine-tuning, I show that open-source LLMs can effectively leverage their pattern recognition capabilities to analyze time series data and astronomical images, tasks that typically demand specialized algorithms and substantial computational resources. I also present a comprehensive framework for developing standardized benchmarks specifically designed for astronomical applications of LLMs.
Short-Bio:
Evgeny Smirnov works in the field of the dynamics of asteroids. His PhD thesis was related to the identification of two-body and three-body mean-motion resonances in the Solar system. In 2017, he introduced a machine-learning approach based on the supervised learning to the identification procedure that decreases the computational time from weeks to seconds. The same year, he proposed a similar approach for asteroid families instead of the classical HCM method. Having a strong background in science and software development, Evgeny connects these areas and brings modern software development patterns and techniques in the field of astronomy.
Google Meet: https://meet.google.com/pcw-gmem-jyi
Link da transmissão: https://www.youtube.com/c/AstronomiaIAGUSP/live