Yonghan (Harry) Yang

I am Yonghan Yang (杨永函), who also goes by Harry — an undergraduate researcher in Artificial Intelligence at the Mohamed bin Zayed University of Artificial Intelligence (MBZUAI) in Abu Dhabi. My work sits at the intersection of generative modeling, agentic AI, and AI for science: I like building methods that are principled enough to prove something about, yet practical enough to run.

Most of my research asks a single question in different disguises — how do we let a model search beyond the data it was trained on without fooling itself? This led to SPADE (ICML 2026), where a calibrated conditional diffusion surrogate is kept honest by a support-proximity prior for offline black-box optimization. I am now extending these ideas to memory retrieval for autonomous agents, discrete diffusion, and biomedical foundation models, in collaboration with researchers at Mila, Harvard Medical School, and GenBio AI.

Before MBZUAI I studied at Beijing National Day School, where I led the school’s synthetic biology and iGEM teams to two consecutive gold medals, and represented China in mathematical modeling. That path — from wet-lab genetic circuits to score-based generative models — is why I care about interdisciplinary work: the most interesting problems rarely respect departmental boundaries.

My motto, borrowed from Cressida Cowell, is to become a hero the hard way.

Research interests

• Generative models — score-based / diffusion models, discrete diffusion, surrogates for optimization
• Agentic AI — memory, retrieval, and evaluation of autonomous LLM agents
• AI for science — AI-driven drug discovery, biomedical ML agents, protein & genetic-circuit modeling
• Scientific ML — physics-informed neural networks, mathematical modeling, and interdisciplinary applications

News

• May 2026 — SPADE accepted to ICML 2026 (co-first author).
• Feb 2026 — Serving as a reviewer for the ICLR 2026 Workshop FM4Science.
• Jan 2026 — SPADE also accepted to the ICLR 2026 DeLTa workshop.
• Aug 2025 — Started my B.Sc. in Artificial Intelligence at MBZUAI on the Sheikh Tahnoon bin Zayed Scholarship in AI Excellence.

Beyond research

When I am not training models, I sketch, play tennis and badminton, run ~70 km a month, and read widely across literature, philosophy, and game theory. I am bilingual in Mandarin and English and am chipping away at French and Japanese. I write about mathematics, learning, and life on the blog.

Feel free to reach out at harryyangyh@outlook.com or connect on LinkedIn and GitHub.