An open 8B model scored 80.9% on an astronomy benchmark. The paper puts that on par with GPT-4o.de Haan et al., Scientific Reports 15 (2025): AstroSage-Llama-3.1-8B, trained on astronomy arXiv papers and synthetic Q&A. Worth Reading, Intermediate. The curator's note: "you don't need a frontier model to achieve acceptable results!"https://readmarginalia.com/reviews/e998f894-947b-4b64-928e-44bbea5495cd/smirik?utm_source=mastodon&utm_medium=social#Astronomy #astrodon #LLM #OpenScience
An open 8B model scored 80.9% on an astronomy benchmark. The paper puts that on par with GPT-4o.de Haan et al., Scientif...
