Jinwoo Ahn

👋 Hey! I am a third-year undergraduate student in EECS at UC Berkeley, working with David Chan on multimodal pragmatic reasoning. I am also a SPAR Research Fellow working with Rishub Jain and Joshua Jacob on trustworthy AI judge systems.

🕐 Previously, I interned twice at Boeing Research, where I worked on computer vision for visual inspection. I also worked with Ruiqi Zhong at BAIR on scalable oversight and natural language processing.

Email  /  Scholar  /  Twitter  /  Github  /  LinkedIn  /  Medium

profile photo

Research Statement

I am interested in understanding how humans and machines "think" differently. For this reason, my work focuses on tasks where AI systems fail to exhibit human-like behavior and reasoning patterns. I believe reducing these gaps would allow humans and AI systems to collaborate more effectively. Besides this, I am also interested in AI for education and AI safety.

Recent News

  • [9/7/2026] I joined SPAR as a Research Fellow.
  • [6/25/2026] I was awarded the JSCM - Joint Service Commendation Medal.
  • [5/24/2026] I officially completed my military service.
  • [6/25/2025] Our VAGUE paper was accepted to ICCV 2025.
  • [1/2/2025] I began my military service at the Combined Forces Command.
  • [9/15/2024] I joined AttentionX as a guest member.
  • [1/15/2024] I returned to Boeing as a Research Intern.
  • [9/21/2023] Our OpenD5 paper was accepted NeurIPS 2023.
  • [7/15/2023] I was accepted to the Berkeley EECS Honors Program.
  • [6/26/2023] I joined Boeing as a Research Intern.
  • [6/12/2022] I joined Dashlabs.ai (YC W21) as a Product Management Intern.

Publications

VAGUE: Visual Contexts Clarify Ambiguous Expressions
Heejeong Nam*, Jinwoo Ahn*, Keummin Ka, Jiwan Chung, Youngjae Yu
International Conference on Computer Vision (ICCV), 2025
[paper] [code] [project]

Goal Driven Discovery of Distributional Differences via Language Descriptions
Ruiqi Zhong, Peter Zhang, Steve Li, Jinwoo Ahn, Dan Klein, Jacob Steinhardt
Advances in Neural Information Processing Systems (NeurIPS), 2023
[paper] [code]

Others

Fine-Grained Open-Vocabulary Object Recognition via User-Guided Segmentation
Jinwoo Ahn, Hyeokjoon Kwon, Hwiyeon Yoo
Technical Report, 2024
[paper]

Recursive Chain-of-Feedback Prevents Performance Degradation from Redundant Prompting
Jinwoo Ahn, Kyuseung Shin
Technical Report, 2024
[paper]

Miscellaneous

  • I come from a multicultural background. I was born in Korea, grew up in China, and am now living in the US.
  • I train mixed martial arts. I occasionally go on amateur leagues and want to fight professionally someday.
  • I have my own way of staying disciplined. I either run, train martial arts, and/or lift at the gym regularly.

Collaboration

I am always open to research collaborations, especially on strong ideas with the potential to develop into strong publications or patents. I frequently pursue independent projects with collaborators outside of my primary affiliation when I come across questions that I find worth investigating. If you are interested in working together, please feel free to reach out.

Below are some of my fantastic collaborators I have had to opportunity to work with external to my primary affiliation:

  • Kyuseung Shin (Everspin; now BS at UC Berkeley)
  • Heejeong Nam (Boeing; now MSCS at Brown)
  • Hwiyeon Yoo, PhD (Boeing; now at Naver Labs)
  • Hyeokjoon Kwon (Boeing; now MSR at CMU)

Website design from Jon Barron