Changliu Liu

Changliu Liu is a roboticist at Carnegie Mellon University who studies trustworthy robotics systems that work with people. She earned engineering and economics degrees at Tsinghua and graduate degrees at UC Berkeley, then completed a postdoctoral appointment at Stanford before joining CMU in 2019. Her work spans human-robot behavior and neural-network verification: she published a book on robot behavior and co-founded the International Neural Network Verification Competition. In 2025, she received the IFAC Robotics Outstanding Young Researcher Award and co-authored BrickGPT, an ICCV Marr Prize paper on stable brick designs from text prompts. As of October 2026, she is an associate professor and leads CMU's Intelligent Control Lab, whose Prompt-to-Product system uses language prompts to create brick assemblies with a bimanual robot.

Now at: Carnegie Mellon University
Works: 234
Citations: 3,146
h-index: 28

Publication numbers from OpenAlex, as of 19 October 2026.

Timeline

2012–2017

Tsinghua and Berkeley degrees

She earns bachelor's degrees in engineering and economics at Tsinghua University in 2012, master's degrees in engineering (2014) and mathematics (2015), and a Ph.D. in engineering at UC Berkeley in 2017, where she works at the Mechanical Systems & Control Lab.
2018

A Stanford postdoc

She works as a postdoctoral researcher at Stanford's Intelligent Systems Laboratory in 2018.
January 2019

Joining Carnegie Mellon

She joins Carnegie Mellon University in January 2019.
2019

A book on robot behavior

She publishes Designing Robot Behavior in Human-Robot Interactions with CRC Press in 2019. The book proposes a framework for human-robot systems and methods for robot control, planning, decision, and learning.
2020

The neural verification competition

She co-founds the International Neural Network Verification Competition in 2020.
2025

An IFAC robotics award

She receives the International Federation of Automatic Control's Robotics Outstanding Young Researcher Award in 2025 for contributions to intelligent robot control with provable safety and human-robot collaboration.
2025

BrickGPT wins the Marr Prize

She co-authors BrickGPT, an autoregressive approach for generating physically stable interconnecting brick assemblies from text prompts. The paper wins the ICCV 2025 Best Paper Award (Marr Prize).
As of October 2026

Leading the Intelligent Control Lab

She is an associate professor in CMU's Robotics Institute and leads the Intelligent Control Lab. Her research focuses on trustworthy robotics systems that work with people. In 2026, her lab's Prompt-to-Product paper generates real-world assemblies from natural-language prompts and uses a bimanual robotic system to build them.