Fenghai Li

About

Hello! My name is Fenghai Li. I am a junior majoring in Computer Science and Mathematics at the University of Illinois Urbana-Champaign (UIUC). I am currently a visiting research intern at the SALT Lab, working with Prof. Diyi Yang's group under the mentorship of Yanzhe Zhang. Previously, I was a research intern at the U Lab with Prof. Jiaxuan You, mentored by Haofei Yu. I have also worked on Human-AI Interaction under the mentorship of Zijian Ding.

I'm always enthusiastic about collaborating with researchers from diverse fields. If you're interested in working together, please don't hesitate to reach out.

Research Interests

My research interests center on Human–AI Collaboration: how language agents can help people work more productively and creatively—not by replacing them, but by working alongside them. I am particularly interested in designing and evaluating Human–AI systems for scientific discovery, portfolio management, and other open-ended tasks.

My work to date has explored these questions in three directions:

Evaluation in live environments. I have evaluated language agents in live, evolving settings with measurable outcomes, using financial markets as a testbed [LiveTradeBench].

Research-agent infrastructure. I have helped build extensible frameworks for creating, customizing, and steering research agents [TinyScientist].

Interfaces for research ideation. I have designed interfaces that help researchers explore trade-offs among ideas [ResearchCube] and directly manipulate and refine underlying concepts [Conceptual DM].

Publications

ResearchCube: Multi-Dimensional Trade-off Exploration for Research Ideation

Zijian Ding, Fenghai Li, Ziyi Wang, and Joel Chan

Under review

LiveTradeBench: Seeking Real-World Alpha with Large Language Models

Haofei Yu, Fenghai Li, and Jiaxuan You

Preprint 2025

TinyScientist: An Interactive, Extensible, and Controllable Framework for Building Research Agents

Haofei Yu*, Keyang Xuan*, Fenghai Li*, Kunlun Zhu, Zijie Lei, Jiaxun Zhang, Ziheng Qi, Kyle Richardson, and Jiaxuan You (*Core Contributors)

The 2025 Conference on Empirical Methods in Natural Language Processing (EMNLP'25), System Demonstrations

Towards Conceptual Direct Manipulation: Drag-Based Research Ideation, Evaluation and Evolution

Zijian Ding, Fenghai Li, Haofei Yu, and Joel Chan

Adjunct Proceedings of the 2025 Annual ACM Symposium on User Interface Software and Technology (UIST'25 Poster), 2025

Open Source Applications

Tiny Scientist

136
Core Contributor

A lightweight framework for building research agents designed for developers.

Live Trade Bench

163
Core Contributor

Live evaluation of AI trading agents in real time across US stocks and Polymarket.