Nick Rui

I'm a math undergrad at Stanford and a researcher with the NLP group at SAIL. Previously, I built software at AWS and caught frogs working in outdoor swimming pool management and operations.

Email  /  LinkedIn  /  X  /  GitHub

Research

I am broadly interested in data, reasoning, and interpretability.

	Where's the Plan? Locating Latent Planning in Language Models with Lightweight Mechanistic Interventions Nicole H. Ma, Nick Rui ICML Workshop on Mechanistic Interpretability, 2026 arXiv We investigate where activation-level representations of future tokens form in open-source language models and how they change with scale.
	Test-Time Meta-Adaptation with Self-Synthesis Zeyneb N. Kaya, Nick Rui ICLR Workshop on AI with Recursive Self-Improvement, 2026 arXiv We introduce a bilevel optimization framework to train language models to self-adapt at inference time using reinforcement learning and meta-gradients.

Math

Awards

Template from Jon Barron.