Reliable · Adaptive · Self-Improving AI

Building robust, adaptive, and self-improving AI systems.

RAISE develops learning algorithms for generative models and agentic AI systems drawing on optimization, probabilistic inference, and representation learning.

Publications

RAISE LAB

About the Lab

The Reliable, Adaptive, and Self-Improving AI Lab (RAISE) develops machine learning methods for AI systems that can adapt to new tasks, learn from feedback, and improve their capabilities reliably over time.

Our research spans three connected areas: machine-learning foundations, including optimization, probabilistic inference, representation learning, and control; generative AI, particularly diffusion and flow-based models; and autonomous, self-improving AI, including test-time scaling, and continual adaptation.

Research Themes

⚙️

Foundations of Machine Learning

Optimization, probabilistic inference, and representation learning, for robust, efficient, and scalable learning systems.

✨

Generative AI

Diffusion, flow-based, and latent-variable models for controllable generation, inverse problems, and scientific applications.

🤖

Autonomous & Self-Improving AI

test-time scaling, and recursive self-improvement through exploration, memory, feedback, and continual adaptation.

Current Members

PhD Students

PhD Student

Loukas Sfountouris

2024 – present

Deep Generative Models

Main supervisor: Paris Giampouras · Co-supervisor: Prof. Theo Damoulas

Website Email

PhD Student

Xietao Wang Li

2024 – present

AI/ML for Science

Co-supervisor: Paris Giampouras · Main supervisor: Prof. Tom Montenegro-Johnson

Website Email