Core AI
– Stochastic Interpolants: Advancing generation, translation, and reconstruction across probabilistic manifolds to seamlessly bridge data distributions with theoretical precision.
– Causality and Marker Identification: Leveraging multimodal data integration to uncover latent mechanisms and interpretable relationships.
– Inverse Problem Solving: Establishing principled frameworks for robust, explainable, and data-efficient model estimation.
– On-Device Intelligent Systems: Enabling adaptive, privacy-preserving learning and real-time inference directly at the edge.
Biomedical AI
– Translational Integration of Core AI: Applying foundational AI methodologies to radiology, pathology, and structural biology to accelerate data-driven precision medicine.
– Generative and Statistical Inference for Complex Molecular Design: Pioneering AI-driven molecular generation and pathway-centric drug discovery through large-scale generative modeling and causal inference.
– Generative and Statistical Inference for Biomedical Multimodality: Advancing multimodal biomedical diagnostics and therapeutics with interpretability and cross-domain reasoning.