Junhyuk Oh

Senior Staff Research Scientist

Google DeepMind

junhyuk (at) google.com

I am a senior staff research scientist at Google DeepMind. My recent work focuses on meta reinforcement learning, LLM post-training, and LLM agents. Before Google DeepMind, I completed my Ph.D. at the University of Michigan, where I was co-advised by Honglak Lee and Satinder Singh.

News

Nov 2025: DiscoRL is now published in Nature. The majority of the work was finished in 2022.

Invited Talks

Jul 2022: ICML 2022 Decision Awareness in Reinforcement Learning, Baltimore, MD.
Jun 2019: RE-WORK Deep Reinforcement Learning Summit, San Francisco, CA.
Jul 2018: ICML 2018 Workshop on Prediction and Generative Modeling in Reinforcement Learning, Stockholm, Sweden. [Slides]
Jan 2018: RE-WORK Deep Learning Summit, San Francisco, CA.
Nov 2017: Ann Arbor Deep Learning Event, Ann Arbor, MI.
Oct 2017: Amazon Graduate Research Symposium, Seattle, WA.
Mar 2017: RE-WORK Machine Intelligence Summit, San Francisco, CA.
Dec 2016: NIPS 2016 Workshop on Deep Reinforcement Learning, Barcelona, Spain. [Slides]
Nov 2016: Ann Arbor Deep Learning Event, Ann Arbor, MI.

Publications

Discovering state-of-the-art reinforcement learning algorithms

Junhyuk Oh, Iurii Kemaev, Greg Farquhar, Dan A Calian, Matteo Hessel, Luisa Zintgraf, Satinder Singh, Hado van Hasselt, David Silver

Nature, 2025

DataRater: Meta-Learned Dataset Curation

Dan A Calian, Gregory Farquhar, Iurii Kemaev, Luisa Zintgraf, Matteo Hessel, Jeremy Shar, Junhyuk Oh, András György, Tom Schaul, Jeffrey Dean, Hado van Hasselt, David Silver

Neural Information Processing Systems (NeurIPS), 2025

Gemini 2.5: Pushing the frontier with advanced reasoning, multimodality, long context, and next generation agentic capabilities

Gemini Team

Tech Report, 2025

Learning from negative feedback, or positive feedback or both

Abbas Abdolmaleki, Bilal Piot, Bobak Shahriari, Jost Tobias Springenberg, Tim Hertweck, Rishabh Joshi, Junhyuk Oh, Michael Bloesch, Thomas Lampe, Nicolas Heess, Jonas Buchli, Martin Riedmiller

International Conference on Learning Representations (ICLR), 2024 (Spotlight)

Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context

Gemini Team

Tech Report, 2024

Gemini: A Family of Highly Capable Multimodal Models

Gemini Team

Tech Report, 2023

Deep Reinforcement Learning with Plasticity Injection

Evgenii Nikishin, Junhyuk Oh, Georg Ostrovski, Clare Lyle, Razvan Pascanu, Will Dabney, André Barreto

Neural Information Processing Systems (NeurIPS), 2023 (Spotlight)

In-context Reinforcement Learning with Algorithm Distillation

Michael Laskin, Luyu Wang, Junhyuk Oh, Emilio Parisotto, Stephen Spencer, Richie Steigerwald, DJ Strouse, Steven Hansen, Angelos Filos, Ethan Brooks, Maxime Gazeau, Himanshu Sahni, Satinder Singh, Volodymyr Mnih

International Conference on Learning Representations (ICLR), 2023 (Oral)

Introducing Symmetries to Black Box Meta Reinforcement Learning

Louis Kirsch, Sebastian Flennerhag, Hado van Hasselt, Abram Friesen, Junhyuk Oh, Yutian Chen

AAAI Conference on Artificial Intelligence (AAAI), 2022

Discovery of Options via Meta-Learned Subgoals

Vivek Veeriah, Tom Zahavy, Matteo Hessel, Zhongwen Xu, Junhyuk Oh, Iurii Kemaev, Hado van Hasselt, David Silver, Satinder Singh

Neural Information Processing Systems (NeurIPS), 2021

Balancing Constraints and Rewards with Meta-Gradient D4PG

Dan A. Calian, Daniel J Mankowitz, Tom Zahavy, Zhongwen Xu, Junhyuk Oh, Nir Levine, Timothy Mann

International Conference on Learning Representations (ICLR), 2021

Pick Your Battles: Interaction Graphs as Population-Level Objectives for Strategic Diversity

Marta Garnelo, Wojciech Marian Czarnecki, Siqi Liu, Dhruva Tirumala, Junhyuk Oh, Gauthier Gidel, Hado van Hasselt, David Balduzzi

International Conference on Autonomous Agents and Multiagent Systems (AAMAS), 2021

Discovering Reinforcement Learning Algorithms

Junhyuk Oh, Matteo Hessel, Wojciech M. Czarnecki, Zhongwen Xu, Hado van Hasselt, Satinder Singh, David Silver

Neural Information Processing Systems (NeurIPS), 2020

Paper Press: VentureBeat Press: Analytics India Press: SingularityHub

Meta-Gradient Reinforcement Learning with an Objective Discovered Online

Zhongwen Xu, Hado van Hasselt, Matteo Hessel, Junhyuk Oh, Satinder Singh, David Silver

Neural Information Processing Systems (NeurIPS), 2020

A Self-Tuning Actor-Critic Algorithm

Tom Zahavy, Zhongwen Xu, Vivek Veeriah, Matteo Hessel, Junhyuk Oh, Hado van Hasselt, David Silver, Satinder Singh

Neural Information Processing Systems (NeurIPS), 2020

What Can Learned Intrinsic Rewards Capture?

Zeyu Zheng, Junhyuk Oh, Matteo Hessel, Zhongwen Xu, Manuel Kroiss, Hado van Hasselt, David Silver, Satinder Singh

International Conference on Machine Learning (ICML), 2020

Grandmaster level in StarCraft II using multi-agent reinforcement learning

Oriol Vinyals, Igor Babuschkin, Wojciech M Czarnecki, Michaël Mathieu, Andrew Dudzik, Junyoung Chung, David H Choi, Richard Powell, Timo Ewalds, Petko Georgiev, Junhyuk Oh, Dan Horgan, Manuel Kroiss, Ivo Danihelka, Aja Huang, Laurent Sifre, Trevor Cai, John P Agapiou, Max Jaderberg, Alexander S Vezhnevets, Rémi Leblond, Tobias Pohlen, Valentin Dalibard, David Budden, Yury Sulsky, James Molloy, Tom L Paine, Caglar Gulcehre, Ziyu Wang, Tobias Pfaff, Yuhuai Wu, Roman Ring, Dani Yogatama, Dario Wünsch, Katrina McKinney, Oliver Smith, Tom Schaul, Timothy Lillicrap, Koray Kavukcuoglu, Demis Hassabis, Chris Apps, David Silver

Nature, 2019

Discovery of Useful Questions as Auxiliary Tasks

Vivek Veeriah, Matteo Hessel, Zhongwen Xu, Richard Lewis, Janarthanan Rajendran, Junhyuk Oh, Hado van Hasselt, David Silver, Satinder Singh

Neural Information Processing Systems (NeurIPS), 2019

Unicorn: Continual Learning with a Universal, Off-policy Agent

Daniel J. Mankowitz, Augustin Žídek, André Barreto, Dan Horgan, Matteo Hessel, John Quan, Junhyuk Oh, Hado van Hasselt, David Silver, Tom Schaul

Reinforcement Learning and Decision Making (RLDM), 2019

Contingency-Aware Exploration in Reinforcement Learning

Jongwook Choi, Yijie Guo, Marcin Moczulski, Junhyuk Oh, Neal Wu, Mohammad Norouzi, Honglak Lee

International Conference on Learning Representations (ICLR), 2019

On Learning Intrinsic Rewards for Policy Gradient Methods

Zeyu Zheng, Junhyuk Oh, Satinder Singh

Neural Information Processing Systems (NeurIPS), 2018

Hierarchical Reinforcement Learning for Zero-shot Generalization with Subtask Dependencies

Sungryull Sohn, Junhyuk Oh, Honglak Lee

Neural Information Processing Systems (NeurIPS), 2018

Many-Goals Reinforcement Learning

Vivek Veeriah, Junhyuk Oh, Satinder Singh

NeurIPS Workshop on Deep Reinforcement Learning, 2018

Generative Adversarial Self-Imitation Learning

Yijie Guo, Junhyuk Oh, Satinder Singh, Honglak Lee

NeurIPS Workshop on Deep Reinforcement Learning, 2018

Self-Imitation Learning

Junhyuk Oh, Yijie Guo, Satinder Singh, Honglak Lee

International Conference on Machine Learning (ICML), 2018

Value Prediction Network

Junhyuk Oh, Satinder Singh, Honglak Lee

Neural Information Processing Systems (NeurIPS), 2017

Zero-Shot Task Generalization with Multi-Task Deep Reinforcement Learning

Junhyuk Oh, Satinder Singh, Honglak Lee, Pushmeet Kohli

International Conference on Machine Learning (ICML), 2017

Control of Memory, Active Perception, and Action in Minecraft

Junhyuk Oh, Valliappa Chockalingam, Satinder Singh, Honglak Lee

International Conference on Machine Learning (ICML), 2016

Paper Code Video Press: MIT Technology Review Press: Daily Mail

Learning Transferrable Knowledge for Semantic Segmentation with Deep Convolutional Neural Network

Seunghoon Hong, Junhyuk Oh, Bohyung Han, Honglak Lee

Computer Vision and Pattern Recognition (CVPR), 2015 (Spotlight)

Paper Code Project

Action-Conditional Video Prediction using Deep Networks in Atari Games

Junhyuk Oh, Xiaoxiao Guo, Honglak Lee, Richard Lewis, Satinder Singh

Neural Information Processing Systems (NeurIPS), 2015 (Spotlight)

Paper Code Video

Experience

Research Scientist

DeepMind

Jun 2018 – Present London, UK

Research Intern

DeepMind

Jun 2017 – Sep 2017 London, UK

Advisor: David Silver

Research Intern

Microsoft Research

Jun 2016 – Sep 2016 Redmond, WA

Advisor: Pushmeet Kohli

Ph.D. in Computer Science and Engineering

University of Michigan

Sep 2014 – May 2018 Ann Arbor, MI

Advisors: Honglak Lee and Satinder Singh

Software Engineer

ESTSoft

Jun 2011 – Jan 2013 Seoul, Korea

Software Engineer

GALA Lab

Jan 2010 – Jun 2011 Seoul, Korea

B.S. in Computer Science and Engineering

Seoul National University

Mar 2007 – Feb 2014 Seoul, Korea

Includes 3 years of mandatory military service.