$ tree ~/research
├── presidential-power/
│ ├── unilateral-actions/
│ └── continuous-cost-constraint-theory/
├── computational-methods/
│ ├── llm-classification/
│ ├── aes/
│ └── multimodal-analysis/
└── political-communication/
└── coming soon
$ whoami
I study the presidency and the separation of powers, with a particular emphasis on measuring the extent to which presidential unilateral actions change the policy status quo. I leverage large language models to analyze the text of unilateral actions to estimate their ideological positions. My methodological interests include applications of LLMs, Bayesian statistics, causal inference with machine learning, and multimodal deep learning analysis of political visual and textual data.
$ cat dissertation.md
A three-paper dissertation developing a unified, content-focused theory of presidential unilateral action. Presidents strategically calibrate the ideological character, magnitude, and electoral targeting of directives to manage institutional constraints and maximize political returns.
Paper 1
Strategic Visibility & Instrument Choice
Presidents strategically pair ideological directives with less visible instruments like memoranda to minimize political costs.
Paper 2
Policy Displacement & Anchored Embedding Scaling
Introduces a measure of policy displacement using AES to locate directives and statutes in a common ideological space.
Paper 3
Swing-State Targeting
Tests whether presidents tailor directives to electorally pivotal constituencies, particularly during election years.
$ ls ~/research/
Work in progress
Politicians may spread conspiracy theories to raise funds from extremists. Uses 2012 and 2016 ANES data. Finding: Among Republicans, conspiracy theory believers contribute more to candidates.
Campaign Finance · Conspiracy Theories · ANES
with Eunseong Oh
Presented at PolMeth 2025
Uses CLIP technology to analyze how political videos portray megadonors. Applies multimodal deep learning to understand the framing of political visual media.
Multimodal Analysis · CLIP · Political Communication
with Karina Alpayeva, Emmanoel Ferreira, Sarah Siddique, and Kim Yi Dionne
Work in progress
Active learning strategies in large enrollment courses. Examines collaborative teaching approaches that improve student outcomes at minority-serving institutions.
Pedagogy · Higher Education · Active Learning
$ tail -f teaching.log
Teaching Assistant — UC Riverside (2023 – Present)
$ open cv.pdf
Large Language Models · Bayesian Statistics · Causal Inference · Machine Learning · Multimodal Analysis · Text Analysis
R / Python