"A Survey of Large Language Models" offers exactly what its name promises: a comprehensive overview of all available open and closed large language models. In addition to facts about training data, used computational resources, used base models, and used methods, the paper offers a detailed explanation of all aspects of large language models. On more than 120 pages, the paper provides a comprehensive overview with a lot of further information on AI training methods, data selection, or prompt engineering. The paper is regularly updated by the participating researchers.

Support our independent, free-access reporting. Any contribution helps and secures our future. Support now:
Bank transfer
Max is managing editor at THE DECODER. As a trained philosopher, he deals with consciousness, AI, and the question of whether machines can really think or just pretend to.
Join our community
Join the DECODER community on Discord, Reddit or Twitter - we can't wait to meet you.