Anthony making salmon nigiri in a kitchen

Hi, I’m Anthony Nguyen, a second-year pure math student at UC Berkeley. I’m currently a SPAR fellow mentored by Jeanne Salle and Sohaib Imran, researching the transferability of interventions on model organisms. Previously, I was a research fellow at Algoverse, mentored by Yuqi Sun and Josh Ying.

I’m broadly interested in AI safety, especially empirical alignment, foundational alignment, and agency. I want to build aligned models from first principles.

Outside of research, I rock climb, lift, hike, host dinners, and read papers and LessWrong. You can find me at anthony.nguyen@berkeley.edu, or at Benchmark Berkeley, the climbing gym.

Attached is my resume.

Selected Papers

  1. SPAR Midterm Report

    Anthony Thien Nguyen, Krzysztof Galias, Carolyn Qian, Anand Shanker, Ahmed Wez, Jeanne Salle†, Sohaib Imran†

    SPAR Fall 2026, October 2026

  2. Sample-Level White-Box Detection of Alignment Faking

    Lakshya Chaudhry*, Tianqin Meng*, Anthony Nguyen*, Yashraj Panwar*, Yuqi Sun†, Zhuofan Josh Ying†

    Mechanistic Interpretability Workshop at ICML 2026

  3. Sample-Level White-Box Auditing of Alignment Faking

    Lakshya Chaudhry*, Tianqin Meng*, Anthony Nguyen*, Yashraj Panwar*, Yuqi Sun†, Zhuofan Josh Ying†

    Second Workshop on Technical AI Governance Research (TAIGR) at ICML 2026

* equal contribution, alphabetical order  ·  † mentor