Hi, I'm
Grace Yuan.

I grew up in Guangzhou, spent my high school years in Melbourne, and now call Cambridge home as a junior at MIT, double-majoring in Mathematics and AI & Decision-Making.

I'm drawn to the inner workings of large language models — my research spans LLM interpretability, multi-agent AI systems, and AI fairness, with work across Harvard Medical School, IBM Research, and MIT CSAIL. On the industry side, I was an AI/ML intern at Microsoft in Redmond, and I'll be joining Tower Research Capital in New York as a Quant Trading Intern next summer.

Before MIT, I represented Australia in international math olympiads including the IMO and EGMO. Outside of work, I'm usually hunting for great foods, reading, or planning my next trip somewhere new.

Grace Yuan

Selected Publications

All papers
Grace Yuan et al. · EACL 2026 HeaLing Workshop (Oral)

Experience

Details
Summer 2026

Quant Trading Intern

Tower Research Capital
Sep 2025 –

Research Intern

Harvard Medical School (Rajpurkar Lab)
May – Aug 2025

AI/ML Intern

Microsoft
Feb – May 2025

Research Intern

IBM Research & MIT CSAIL (Torralba Lab)
Sep 2024 – May 2025

SERC Research Scholar

MIT Schwarzman College of Computing

Awards

All awards
EGMO Gold Medal (6th globally)
IMO Honourable Mention + Maryam Mirzakhani Prize
Putnam Top 500

Projects

All projects
A robot that sees, thinks, and throws — using GPT-4o for zero-shot waste classification and ballistic tossing. Python Drake Robotics
Model-agnostic method using KG embeddings to fix bidirectional reasoning in LLMs. Python PyTorch
Evaluating FinBERT vs BERT for financial sentiment via UMAP visualization. Python UMAP