Research
Explore our research areas and featured projects.
Our research develops a data-centric foundation for machine learning—treating datasets not as passive artifacts but as structured, dynamic objects that can be characterized, transformed, and optimized. We draw on tools from optimal transport, information theory, and geometric deep learning to formalize how data properties affect learning outcomes.
Dataset Characterization & Geometry
Understanding what makes data valuable for learning through geometry and optimal transport
Dataset Transformations & Training Dynamics
How data structure affects learning dynamics and model behavior during training
Dataset Optimization & Synthesis
Principled methods to reduce, enhance, and synthesize training data
Adaptive & Reconfigurable Models
Dynamically combining and adapting models based on constraints and objectives