Story

This article discusses the progress of large language models over the past three years, from the release of GPT-3 to the development of the Gemini model, which has improved capabilities in areas such as commonsense reasoning and language generation.