About
I am a Master's student in Machine Learning at Carnegie Mellon University. I received my B.S. in Computer Engineering with Highest Honors (and a minor in Mathematics) from the University of Illinois at Urbana-Champaign (UIUC), where I was fortunate to work with Prof. Jiaxuan You on multimodal AI and temporal reasoning in LLMs.
My research interests span machine learning, multimodal AI systems, and large language models. I have contributed to projects on enabling text-only LLMs to reason about visual information (SeeingEye) and improving temporal reasoning capabilities of LLMs (Time-R1).
Previously, I interned at Skytech as an Algorithm Engineer working on LLM prompt engineering, and conducted research on ML-driven printing optimization at the UIUC DIAO Group.
Outside of research, I am originally from Nanjing, China. I enjoy all things cute -- especially Kuromi! I'm an avid player of Monster Hunter and CS2, love creating with Blender, and play badminton in my free time. My dream is to eat my way around the world.
Selected Publications
View All →SeeingEye: Agentic Information Flow Unlocks Multimodal Reasoning In Text-only LLMs
Weijia Zhang†, Zijia Liu†, Haoru Li†, Haoqi Chen†, Jiaxuan You
arXiv preprint
An agent-based framework that enables text-only LLMs to perform multimodal reasoning through agentic information flow architecture.
Time-R1: Towards Comprehensive Temporal Reasoning in LLMs
Zijia Liu, Peixuan Han, Haofei Yu, Haoru Li, Jiaxuan You
arXiv preprint
A comprehensive evaluation and training framework for improving temporal reasoning capabilities in large language models.
An Artificial Intelligence Based Method for Bartending
Haoru Li
IEEE/CIC International Conference on Communications in China (ICCC)
A novel AI-based approach for automated bartending, accepted by IEEE/CIC ICCC 2025.
News
Released [*agent-harness*](https://github.com/Effervescebt/agent-harness), a runtime for tool-using LLM agents, with experiments on how context compression interacts with prompt caching and on self-hosted serving with vLLM.
Released [*Internship Autopilot*](https://github.com/Effervescebt/internship-autopilot), an open-source pipeline that screens internship postings for visa eligibility and tailors a one-page resume to each role.
Started my Master's in Machine Learning at **Carnegie Mellon University**!
Graduated from UIUC with a B.S. in Computer Engineering, with **Highest Honors** and a minor in Mathematics.
Our paper [*SeeingEye*](https://arxiv.org/abs/2510.25092) is available on arXiv.
Our paper [*Time-R1*](https://arxiv.org/abs/2505.13508) is available on arXiv.
Paper [*An Artificial Intelligence Based Method for Bartending*](https://ieeexplore.ieee.org/document/10942013) accepted by IEEE/CIC ICCC 2025.
Awarded the Yunni and Maxine Pao Memorial Scholarship and Gary Whitledge Scholarship.