Research

Work in progress.

I'm an EECS student at Berkeley doing machine-learning research. Most of it is reinforcement learning and how people make the same decision over and over, with some generative modeling for protein design on the side. What I care about most is whether a result still holds once you push on it.

Google Scholar

Publications

In preparation · Targeting ACM CHI 2027 · 2026

Reward Salience and Learning in Repeated Human Decisionsworking title

Nicholas Chen (first author), with Prof. Sinchaisri

A behavioral study of a repeated bundling task across 85 participants and 1,000+ rounds. Fixed-effects causal inference finds that increasing payout salience raised suboptimal choices by 71 percentage points, with no measurable learning over 15 rounds. Behavior separates into five statistically distinct decision policies (8× outcome spread), analyzed with offline policy evaluation over a DQN trained with reward shaping, experience replay, and target networks.

Published · International Journal of High School Research, 7(4) · 2025

ML-Driven Approach to Discovery of Peptide-Based BACE1 Inhibitors

Nicholas Chen · Mentor: Vladimir Akhmetov

A machine-learning approach to the de novo design of catalytic-site-targeted peptide binders of β-secretase (BACE1) — a key enzyme in amyloid-beta plaque formation in Alzheimer's disease. Applies deep reinforcement learning and diffusion models (AlphaFold3, RoseTTAFold) to generate potent, low-immunogenicity inhibitor candidates.

PDF

Appointments

Sep 2025 — Present

AI/ML & Quantitative Research

UC Berkeley — Operations & Behavioral Analytics Lab

Reinforcement learning, offline policy evaluation, and fixed-effects causal inference over a 1,000-round human-decision experiment, with the full experiment and data pipeline in SvelteKit/Firebase.

Feb 2023 — Apr 2024

Machine Learning Researcher

MIT — Pentelute Lab

Built a deep generative and reinforcement-learning system for molecular design over a multi-objective reward; first-authored a peer-reviewed paper.

Dec 2023 — Jan 2024

Computational Research — Ranked 1st of 45

University of Oxford

Monte Carlo, stochastic-process, and time-series models of nonlinear biochemical systems, with PCA, k-means, and random-forest pipelines.

Talks

2026

Reward salience and decision-making in repeated human choice

Presented at Stanford University and Carnegie Mellon University