Diederik KingmaDiederik Kingma

Dutch computer scientist. PhD cum laude from the University of Amsterdam (2017) under Max Welling. Created two of the most fundamental building blocks of modern AI: the Variational Autoencoder (VAE, 2013) -- bridging probabilistic modeling and deep learning, enabling generative models to learn latent structure -- and the Adam optimizer (2014, with Jimmy Ba) -- the most widely used optimization algorithm in deep learning with 95,000+ citations. Part of the founding team of OpenAI (2015). Worked at Google Brain on large-scale optimization and generative models. Joined Anthropic in October 2024. Won the ICLR Test of Time Award in 2024 (VAE) and 2025 (Adam) -- two consecutive years.