Machine Intelligence Lab

Machine Intelligence Lab · Led by Hongki Lim

Generative modeling
and probabilistic inference

We develop learning and inference methods that combine data, prior knowledge, and constraints to solve problems under uncertainty.

Explore our research
Learning → conditioning
Conditioning a learned distribution The figure illustrates a learned prior distribution and its conditional distribution. Dark samples show the prior. A blue band marks a noisy observation; red conditional samples narrow the possibilities while retaining both modes. SAMPLE SPACEx₁x₂ OBSERVATION
Learned distributionConditional samples
Conditioning narrows the possibilities while retaining multiple solutions.
01
Generative modelingLearn distributions
02
Conditional inferenceCondition on evidence
03
Inverse designGenerate under constraints

News

Show more newsShow less news

Join the lab

We welcome students and postdocs interested in generative models, probabilistic inference, and inverse problems.

Research with us