haoran.
Menu +

COMPUTER SCIENCE · HUST

Haoran Qian 钱皓冉

Learning to model the world,
one question at a time.

World Models · Generative Models · Video Generation
Multimodal Learning · Embodied Intelligence

钱皓冉手持相机与拍立得照片的生活留影
把此刻,留给以后。

日子很慢,快门很轻。

UPDATES

News

Milestones from research, study, and the everyday.

🎓 Joined Huazhong University of Science and Technology

Began studying Computer Science at HUST in Wuhan.

🤖 Joined the AI Group at Unique Studio

Started learning and building with the AI group at HUST’s Unique Studio.

🏆 National First Prize at RAICOM

Our team received a National First Prize in the RAICOM Robotics Developer Competition.

01 /

Publications

All publications ↗
DreamTamer: From Lucky Samples to Reliable Futures in Action-Conditioned World Models method overview

DreamTamer: From Lucky Samples to Reliable Futures in Action-Conditioned World Models

Haobo Li, Yanhong Zeng, Zeyu Lou, ZePeng Lin, Haoran Qian, Qin Zhao, Zhihao Xie, Zhipeng Zhang.

An action-conditioned world model trained with reinforcement learning to keep embodied simulation stable.

ICLR 2027 (In Submission)

02 /

Projects

Project index ↗

Things made to understand things.

Implementations, evaluation tools, reproductions, and side projects—documented through the artifacts they produce.

03 /

Experience

Selected record ↗

Research Experience

2026.6–Present
Shanghai Jiao Tong University emblem

Shanghai, China

Research

Shanghai Jiao Tong University

Advised by Zhipeng Zhang.

Education

2025—Present
Huazhong University of Science and Technology emblem

Wuhan, China

Computer Science Undergraduate

Huazhong University of Science and Technology

Undergraduate study in Computer Science at Huazhong University of Science and Technology.

Computer Science
04 /

Fragments

把走过的风景,
折成日子里的一页。

Enter the archive ↗