Foundation Model Lab

Master cutting-edge AI by fine-tuning models for material science (with Microsoft’s MatterGen), drug discovery, and more. Outcome: Become a leading AI Engineer, Foundation Model Engineer, or Data Engineer.

Welcome to the Foundation Model Lab

where cutting-edge AI research meets real-world application. We are at the frontier of a new computational era — pioneering the fine-tuning, alignment, and deployment of large-scale AI models for scientific discovery, material innovation, and personalized healthcare. From Microsoft’s MatterGen to open-source giants like LLaMA and Mistral, our lab dives deep into the core of what makes modern AI truly intelligent.

 Our mission is bold to create domain-specific foundation models that will power the next decade of innovation. As the demand for smarter, more adaptive AI grows exponentially, we are not simply adopting the future — we are building it. 

For Students Become a Foundation Model Architect

Do you aspire to move beyond the traditional data scientist role? At the Foundation Model Lab, you don’t just learn how models work — you engineer them from the ground up. This is where undergraduates work on problems typically tackled by AI researchers at the world’s top labs.

What We Offer:

01 A Visionary Three-Year Program

Structured into six high-intensity levels (each six months long), our program takes you from fundamentals to frontiers. Students work in pods, tackling challenges across biology, chemistry, physics, and computation.

02 Impactful Real-World Projects

Work with industry collaborators like Microsoft, OpenAI ecosystem contributors, and biotech startups. Build models that help discover new drugs, invent new materials, and simulate real-world systems with precision.

03 A Culture of Deep Scientific Exploration

Welcome to a 9-to-9 lab environment — a space that echoes with ideas, algorithms, and ambition. Here, you’ll engage in original research, weekly paper readings, and hackathons, alongside AI mentors and domain experts.

A Three-Year Journey

Navigating Six Levels of Mastery

The six levels are strategically divided across the three years, with two levels to be conquered each year. This pacing allows for in-depth learning and practical application at each stage.

Level 1

Remembering

Focuses on recalling key facts, terms, and basic AI concepts. Students memorize, list, identify principles of foundation models, answering “what,” “where,” and “when” questions.

Level 2

Understanding

Involves explaining and summarizing learned concepts, comparing models (e.g., LLaMA vs. Mistral), and answering “how” and “why” to show deeper grasp of foundation models.

Level 3

Applying

Students use their knowledge in real-world tasks like fine-tuning models or applying tools (e.g., MatterGen), bridging theory and practice.

Level 4

Analyzing

Focuses on breaking down AI models, identifying patterns, and comparing architectures to develop critical understanding and evaluation skills.

Level 5

Evaluating

Students critically assess AI models, judge credibility and relevance, and justify their opinions with logic and evidence.

Level 6

Creating

Students synthesize knowledge to design original AI solutions, models, or research, showcasing innovation and expertise in foundation models.

Our Collaborative Ecosystem A Constellation of Labs

The Foundation Model Lab sits at the computational core of an interdisciplinary galaxy of research environments. You’ll work shoulder-to-shoulder with experts across these labs:

Embodied AI Brain Lab

Integrate our multi-agentic brains into the physical world. Work on projects involving robots, exoskeletons, surgical bots, and other hardware to give intelligence a physical form.