About

Hello, I'm Haosheng Wang, a Master's student in Computer Science at Brown University. I previously earned a joint degree in Mathematics and Computer Science from New York University. I am broadly interested in artificial intelligence, with a long-term goal of developing intelligent systems that can reason, learn, and act in grounded and generalizable ways—and ultimately contribute to expanding human knowledge.

I am particularly drawn to questions of abstraction, memory, and adaptation: how intelligent systems can operate efficiently in high-dimensional environments, accumulate and reuse knowledge over time, and balance neural representations with symbolic structure. I view AI not only as a powerful engineering discipline, but also as a lens for understanding the principles underlying human intelligence itself.

Research Interests

Long-term memory and efficient adaptation
How can intelligent systems build and use long-term memory—internally or via external structured representations—to support robust generalization and domain adaptation?
LLM adaptation generalization continual learning neural–symbolic memory
Goal design and self-evaluation
Humans learn without explicit labels or reward functions. How can AI agents develop self-evaluation signals and shape their own objectives to improve over time?
self-supervised learning preference learning self-improvement goal / utility design
Abstraction: symbols and embeddings
What are the tradeoffs between discrete symbols and continuous embeddings for efficiency, interpretability, and scalability—and how can hierarchical abstraction bridge them?
neural–symbolic systems abstraction hierarchical RL representation learning

Education

Brown University — M.S., Computer Science
Providence, RI · Sep 2024 -- May 2026 · Focus: AI/ML
New York University — B.A., Math and Computer Science
New York, NY · Jan 2021 -- May 2024

Contact

Best way to reach me: hswang947@gmail.com. You can also find me on LinkedIn.