I’m a 3rd year PhD Student at USC, working with Dani Yogatama and Yan Liu. Currently, I’m visiting Harvard, working with David Alvarez-Melis and Naomi Saphra.
I’m interested in interpretability - how we make sense of models, and how it might uncover the underlying science of large-scale models. In particular, I focus on understanding training and reasoning, as well as how these insights can be applied in useful, practical, and actionable ways.
Static evaluations are very brittle—especially concerning for extremely high-steaks medical settings like organ allocations. Turns out, simply asking LLMs to rank recipients (which is what actual organ transplant committees do) breaks apparent LLM fairness.
We introduce a large-scale, complexity-annotated, verified CoT-like reasoning trace dataset showing that current LLMs still struggle with structured inference.
