|
Sanghyun Yi I'm a Senior Applied Scientist at a stealth AI startup, working on agentic AI for higly regulated industries. I'm also interested in bridging biological and artificial intelligence: understanding human decision-making with machine learning, and solving ML problems from neuroscience. I received my Ph.D. in Social and Decision Neuroscience at Caltech, advised by John P. O'Doherty. In my free time, I love going to the gym, collecting model trains, and working on my vintage car :) |
|
Publications |
|
Computational and neural mechanisms underlying the influence of action affordances on value learning
Sanghyun Yi and John P. O'Doherty Nature Communications, 2026 |
|
Hardware-Friendly Static Quantization Method for Video Diffusion Transformers
Sanghyun Yi, Qingfeng Liu, and Mostafa El-Khamy arXiv, 2025 arXiv |
|
Neural mechanisms underlying the hierarchical construction of perceived aesthetic value
Kiyohito Iigaya, Sanghyun Yi, Iman A. Wahle, Koranis Tanwisuth, Logan Cross, and John P. O'Doherty Nature Communications, 14 (1), 127, 2023 |
|
Aesthetic preference for art can be predicted from a mixture of low- and high-level visual features
Kiyohito Iigaya, Sanghyun Yi, Iman A. Wahle, Koranis Tanwisuth, and John P. O'Doherty Nature Human Behaviour, 5, 743–755, 2021 PDF / Science |
|
Towards Coherent and Engaging Spoken Dialog Response Generation Using Automatic Conversation Evaluators
Sanghyun Yi, Rahul Goel, Chandra Khatri, Alessandra Cervone, Tagyoung Chung, Behnam Hedayatnia, Anu Venkatesh, Raefer Gabriel, and Dilek Hakkani-Tur International Conference on Natural Language Generation (INLG), Tokyo, Japan, 2019 (Oral) |
|
Science meets art: Attribute-based computation as a general principle for building subjective value
Kiyohito Iigaya, Sanghyun Yi, Iman Wahle, Sandy Tanwisuth, and John P. O'Doherty Multi-disciplinary Conference on Reinforcement Learning and Decision Making (RLDM), Montréal, Canada, 2019 |
|
A Competitive Path to Build Artificial Football Agents for AI Worldcup
Sanghyun Yi, Jeehang Lee, Changhwa Lee, Juno Kim, Sujin An, and Sangwan Lee IEEE/IEIE International Conference on Consumer Electronics (ICCE) Asia, Jeju, Korea, 2018 |
|
Maximally separating and correlating model-based and model-free reinforcement learning
Sanghyun Yi, Jeehang Lee, and Sangwan Lee Computational and Systems Neuroscience (Cosyne), Denver, USA, 2018 |
|
A Chatbot by Combining Finite State Machine, Information Retrieval, and Bot-Initiative Strategy
Sanghyun Yi and Kyomin Jung 1st Proceedings of Alexa Prize (Alexa Prize 2017), Las Vegas, USA, 2017 |
Miscellanea |
Experience |
Senior Applied Scientist, May 2025 – Present Stealth AI startup Applied Scientist Intern (Applied Scientist II), Sep 2024 – Jan 2025 Amazon Web Services, Amazon AI Systems Research Scientist Intern, Jun 2024 – Sep 2024 AI Systems, Samsung Semiconductor, Inc. Applied Scientist Intern, Mar 2018 – Aug 2018 Alexa AI, Amazon Lab126 Undergraduate Research Intern, Jun 2017 – Feb 2018 Laboratory for Brain and Machine Intelligence, KAIST Undergraduate Research Intern, Jan 2016 – Dec 2016 Machine Intelligence Lab, Seoul National University Company Commander (Sergeant), Oct 2013 – Jul 2015 259 Company, Gangbuk Police, Seoul Metropolitan Police Agency Undergraduate Research Intern, Jun 2011 – Aug 2011 Natural Language Processing Group, MIT |
Education |
Ph.D., Social and Decision Neuroscience Caltech, Sep 2018 – Sep 2024 Chen Graduate Fellow B.S. in Mathematics (Minor in Statistics and Computer Science) Seoul National University, 2011 – 2018 (2013 – 2015 in military service) |
Awards |
Chen Graduate Innovator Grant Awards, 2022 ($10,000)
Chen Graduate Fellow, Caltech, 2018 – 2019 NSF Graduate Research Fellowship, 2020 (Honorable Mention) National Science and Technology Scholarship, 2011 – 2017 (Full tuition & fee) Semifinalist, The Alexa Prize, 2016 – 2017 (10th place) 3rd place, Nvidia Deep Learning Contest, 2016 3rd place, AI World Cup 2017, 2017 |
Patents |
Mostafa El-Khamy, Qingfeng Liu, Sanghyun Yi, "Post-training quantization for diffusion transformers", US patent pending (19258874)
Rahul Goel, Chandra Prakash Khatri, Tagyoung Chung, Raefer Christopher Gabriel, Anushree Venkatesh, Behnam Hedayatina, Sanghyun Yi, "Dialog Response Generation", US patent 11194973 Sang Wan Lee, JeeHang Lee, Sanghyun Yi, "Apparatus and method for eliciting optimal strategy of the humans in the interactive games using artificial intelligence", US patent 11626030 Sang Wan Lee, JeeHang Lee, Sanghyun Yi, "Apparatus and method for eliciting optimal strategy of the humans in the interactive games using artificial intelligence", Korean patent 1021118570000 |
Teaching |
TA for Psy 13: Introduction to Cognitive Neuroscience, Spring 2020, 2021, 2022, 2023
TA for EC/ACM/CS 112: Bayesian Statistics, Winter 2020, 2021, 2022 |
Skills |
Python, PyTorch, LLM, RAG, Agentic AI, MCP, RL, SFT, NLP, Deep Learning |
|
Website template from Jon Barron. |