Jacob Gardner

Jacob Gardner is an Assistant Professor in the Department of Computer and Information Science at the University of Pennsylvania, where he leads a research group focused on probabilistic machine learning. His work bridges theoretical foundations and practical applications, with emphasis on Bayesian optimization and generative modeling for scientific challenges like drug discovery and materials design.

Education & Background

He earned his Ph.D. in Computer Science from Cornell University under Kilian Weinberger, followed by postdoctoral work in Cornell's Operations Research department. Prior to joining Penn, he was a Research Scientist at Uber AI Labs.

Research Focus

His lab develops methods for:

• Bayesian optimization for high-dimensional scientific design problems
• Scalable Gaussian process inference
• Black-box variational inference with convergence guarantees
• Adversarial robustness in machine learning systems

Publication Trends

Recent work (2022-2023) demonstrates strong focus on efficient Bayesian methods: 80% of publications center on optimization techniques (local/global Bayesian optimization) and variational inference, with applications to latent space modeling and hyperparameter tuning. Papers consistently appear at NeurIPS, ICML, and AISTATS.

Research Team

Advises five PhD students:

• Natalie Maus (Bayesian optimization)
• Kaiwen Wu (optimization theory)
• Kyurae Kim (variational inference)
• Haydn Jones (structured data)
• Yimeng Zeng (generative modeling)

Software Contributions

Co-founded GPyTorch, a PyTorch-based framework for GPU-accelerated Gaussian processes that replaces traditional Cholesky decomposition with linear conjugate gradient methods.