Jimmy Ba

Papers

Total Citations

H-Index

About

Jimmy Ba is a prominent machine learning researcher whose work spans reinforcement learning, world models, and automated system design. Best known for foundational contributions to deep learning optimization — including the widely adopted Adam optimizer — Ba has built a research program that pushes the boundaries of how intelligent agents learn and adapt with minimal human intervention. His paper "Mastering Atari with Discrete World Models" (2020, 23 citations) exemplifies his commitment to sample-efficient reinforcement learning, demonstrating how agents can learn robust behaviors by reasoning through imagined futures derived from compact world representations. This work advances a critical challenge in AI: enabling agents to generalize meaningfully from limited experience in visually complex environments. Complementing this, "Neural Graph Evolution: Towards Efficient Automatic Robot Design" (2019, 33 citations) tackles the ambitious problem of removing human engineering from robot design pipelines entirely, navigating vast combinatorial spaces to discover effective locomotion structures automatically. Together, these works reflect Ba's broader mission: reducing the human effort required to build intelligent, adaptive systems. His research is particularly influential among students exploring the intersection of optimization, model-based reinforcement learning, and autonomous design, making him a essential figure in modern deep learning.