Princeton · computer science · class of 2028
Helen Hui
Helen Hui is a Princeton University computer science undergraduate, class of 2028, researching continual learning, hardware-aware model design, scaling, and efficient inference.
Research
- Distillation on reasoning models — Princeton Language and Intelligence Lab, under Sanjeev Arora
- Scaling laws of multi-agent systems — Tom Griffiths lab
Interests: Continual learning, Hardware-aware model design, Scaling laws, Efficient inference, Distillation of reasoning models, Multi-agent systems.
Projects & art
Research & technical writing
Essays
Find Helen
Profile API
This public, read-only endpoint returns the profile, education, research, projects, essays, and official profile links as JSON. No API key is required.
GET https://ilovehhhyn.github.io/helenhui/info/profile.json
curl -fsSL 'https://ilovehhhyn.github.io/helenhui/info/profile.json'
Schema version: 1.0. Updated: .
View the complete JSON response
{
"schema_version": "1.0",
"updated_at": "2026-09-19",
"url": "https://ilovehhhyn.github.io/helenhui/info/",
"api_url": "https://ilovehhhyn.github.io/helenhui/info/profile.json",
"source": {
"url": "https://ilovehhhyn.github.io/helenhui/",
"description": "Self-described public information from Helen Hui’s personal website. Research descriptions reflect the homepage at the time of this update."
},
"person": {
"@context": "https://schema.org",
"@type": "Person",
"@id": "https://ilovehhhyn.github.io/helenhui/#helen-hui",
"name": "Helen Hui",
"url": "https://ilovehhhyn.github.io/helenhui/",
"mainEntityOfPage": "https://ilovehhhyn.github.io/helenhui/info/",
"description": "Helen Hui is a Princeton University computer science undergraduate, class of 2028, researching continual learning, hardware-aware model design, scaling, and efficient inference.",
"jobTitle": "Computer science undergraduate",
"affiliation": {
"@type": "CollegeOrUniversity",
"name": "Princeton University",
"url": "https://www.princeton.edu/"
},
"knowsAbout": [
"Continual learning",
"Hardware-aware model design",
"Scaling laws",
"Efficient inference",
"Distillation of reasoning models",
"Multi-agent systems"
],
"sameAs": [
"https://github.com/ilovehhhyn",
"https://www.linkedin.com/in/helen-hui-7125b929b/",
"https://x.com/helentomato0o"
]
},
"education": {
"institution": "Princeton University",
"field": "Computer science",
"level": "Undergraduate",
"class_year": 2028
},
"research": [
{
"topic": "Distillation on reasoning models",
"lab": "Princeton Language and Intelligence Lab",
"advisor": "Sanjeev Arora",
"source_url": "https://ilovehhhyn.github.io/helenhui/"
},
{
"topic": "Scaling laws of multi-agent systems",
"lab": "Tom Griffiths lab",
"source_url": "https://ilovehhhyn.github.io/helenhui/"
}
],
"projects": [
{
"name": "i draw",
"url": "https://ilovehhhyn.github.io/helenhui/gallery.html"
},
{
"name": "docs agent: choobi",
"url": "https://github.com/ilovehhhyn/choobi"
},
{
"name": "brain dump app: dooni",
"url": "https://ilovehhhyn.github.io/dooni/"
},
{
"name": "drinking pet: tiddly",
"url": "https://ilovehhhyn.github.io/tiddly/"
},
{
"name": "poetry",
"url": "https://ilovehhhyn.github.io/theresNoEnglishWordForXingFu/"
},
{
"name": "gluh gluh gulp",
"url": "https://guulp.vercel.app"
}
],
"research_and_writing": [
{
"name": "helen's takes on hardware aware continual learning",
"url": "https://ilovehhhyn.github.io/helenhui/research-hardware-aware-continual-learning.html"
},
{
"name": "blog: the intuition behind KL divergence in post-training",
"url": "https://ilovehhhyn.github.io/helenhui/research-kl-divergence.html"
},
{
"name": "blog: Why LLMs today can never think like humans",
"url": "https://ilovehhhyn.github.io/helenhui/blog-why-llms-cannot-think-like-humans.html"
},
{
"name": "KL recovery law project repo",
"url": "https://github.com/ilovehhhyn/spin-me-around/tree/main"
},
{
"name": "proposal preprint",
"url": "https://ilovehhhyn.github.io/helenhui/research-forgetting-law-proposal.pdf"
},
{
"name": "project: memory retrival system benchmark",
"url": "https://ilovehhhyn.github.io/helenhui/research-memory-benchmark.html"
},
{
"name": "project: rl-inspired portfolio optimization",
"url": "https://ilovehhhyn.github.io/helenhui/research-portfolio-optimization.html"
}
],
"essays": [
{
"name": "Evolution of thought on writing",
"url": "https://ilovehhhyn.github.io/helenhui/page1.html"
},
{
"name": "On time",
"url": "https://ilovehhhyn.github.io/helenhui/page2.html"
},
{
"name": "On festivities and utility",
"url": "https://ilovehhhyn.github.io/helenhui/page3.html"
},
{
"name": "On love at 18 years old",
"url": "https://ilovehhhyn.github.io/helenhui/page4.html"
},
{
"name": "On starfish",
"url": "https://ilovehhhyn.github.io/helenhui/page5.html"
},
{
"name": "On gratitude",
"url": "https://ilovehhhyn.github.io/helenhui/page6.html"
}
],
"api": {
"method": "GET",
"authentication": "None",
"format": "application/json",
"delivery": "Static JSON hosted on GitHub Pages; no server or query parameters required."
}
}Source: Helen’s personal website. Research descriptions reflect the homepage at the time of this update.