Yunong Liu is a research engineer at Luma AI, where she led Layering, built experimental post-training workflows for Ray3 and contributed to Uni-1. She earned computer science and engineering degrees at Stanford and Edinburgh and led Stanford’s IKEA Manuals at Work research on grounding assembly videos in 3D models and instructions.

Unclaimed profile

Yunong Liu is a research engineer at Luma AI, where she led Layering, built experimental post-training workflows for Ray3 and contributed to Uni-1. She earned computer science and engineering degrees at Stanford and Edinburgh and led Stanford’s IKEA Manuals at Work research on grounding assembly videos in 3D models and instructions.

Career

Research Scientist at NVIDIA from 2026-09 to Present
Research Engineer at Luma AI from 2025-04 to Present
Research Assistant, Vision and Learning Lab at Stanford University from 2023-06 to 2025-04
Teaching Demonstrator at University of Edinburgh from 2022-09 to 2023-01
Deep Learning Intern at Zealen AI from 2021-08 to 2021-09
Student Tech Intern at NXP Semiconductors from 2021-05 to 2021-08
yunongliu1.github.io

IKEA Manuals at Work

First-author research project aligning assembly instructions, internet videos and 3D furniture models to study grounded visual understanding.

Open ↗
arxiv.org

CaptionQA: Is Your Caption as Useful as the Image Itself?

Shijia Yang, Yunong Liu, Bohan Zhai and collaborators introduce a benchmark that tests whether image captions preserve the information needed to answer questions across four application domains.

Open ↗

Yunong's Collections

Yunong's Lists

In these lists

Cosigns Given