GPT-4 Has 1.8 Trillion Parameters. It Uses 2% of Them Per Token.

Author(s): DrSwarnenduAI Originally published on Towards AI. GPT-4 Has 1.8 Trillion Parameters. It Uses 2% of Them Per Token. DeepSeek-R1: 671 billion parameters. 37 billion active per token. DeepSeek-R1: 671 billion parameters. 37 billion active per token.The article discusses various machine learning models, focusing on their parameter count and operational efficiencies. It delves into the architecture of the Mixture of Experts (MoE), detailing how different models utilize parameters per token and how routing affects performance, emphasizing the benefits of utilizing multiple experts for token processing to improve training stability and efficiency. Additionally, it explores specific implementations, such as the DeepSeek model, and compares it with existing architectures to elucidate advantages in computation and memory usage. Read the full blog for free on Medium. Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming a sponsor. Published via Towards AI