PeopleofSuperintelligence

Researcher

Jonathan J. Hunt

Also known as Jonathan Hunt, JJ Hunt

A reinforcement learning researcher, Jonathan J. Hunt coauthored the 2015 paper on continuous control with deep reinforcement learning, which introduced the DDPG algorithm and received an ICLR 2026 Test of Time award. With a computational neuroscience PhD from the University of Queensland in 2011, he joined DeepMind in 2014 and coauthored the 2016 successor features paper on transfer in reinforcement learning. At Twitter, he coauthored Hyperbolic Deep Reinforcement Learning in 2022, which ICLR 2023 presented as a top 25% paper. Having cofounded Elefant AI and led its research team, which open-sourced a vision-language-action model for 3D game playing and 8000 hours of human-annotated gameplay videos, he is a research scientist at Google DeepMind, his second stint at the lab.

Badges

Timeline

  1. 2006–2011

    Physics and neuroscience degrees

    He earns a BSc Hons in mathematical physics from Massey University in New Zealand in 2006. He completes a PhD in computational neuroscience at the University of Queensland in Australia in 2011.

  2. 2013

    Sparse coding and visual cortex

    At the Queensland Brain Institute of the University of Queensland, he coauthors a 2013 paper in PLoS Computational Biology with Peter Dayan and Geoffrey J. Goodhill. The work applies three sparse coding models to binocular receptive field development across six abnormal rearing conditions and matches the receptive field changes seen experimentally.

  3. 2014–2015

    DeepMind and the DDPG paper

    He joins DeepMind in 2014. In 2015, he coauthors Continuous control with deep reinforcement learning with Timothy P. Lillicrap, Alexander Pritzel, Nicolas Heess, Tom Erez, Yuval Tassa, David Silver and Daan Wierstra, presenting an actor-critic, model-free algorithm based on the deterministic policy gradient that solves more than 20 simulated physics tasks.

  4. 2016

    Successor features for transfer

    In 2016, he coauthors Successor Features for Transfer in Reinforcement Learning with André Barreto, Will Dabney, Rémi Munos, Tom Schaul, Hado van Hasselt and David Silver. The paper proposes a transfer framework for tasks that share the same environment dynamics and have different reward functions.

  5. 2022

    Research at Twitter

    His first stint at DeepMind ends in 2019, and he works at Twitter. In 2022, he coauthors Hyperbolic Deep Reinforcement Learning with Edoardo Cetin, Benjamin Chamberlain and Michael Bronstein, which models latent representations in hyperbolic space for deep reinforcement learning algorithms, and ICLR 2023 presents it as a top 25% paper.

  6. January 2026

    Elefant AI and open-p2p

    He cofounds Elefant AI and leads its research team, which open-sources a vision-language-action model for general 3D game playing together with 8000 hours of human-annotated gameplay videos. In January 2026, he coauthors a paper on scaling behavior cloning that releases 8300+ hours of human gameplay, training and inference code, and pretrained checkpoints under an open license.

  7. 2026

    ICLR Test of Time Award

    ICLR 2026 selects Continuous control with deep reinforcement learning, by Lillicrap, Hunt, Pritzel, Heess, Erez, Tassa, Silver and Wierstra, as one of two Test of Time winners among the papers published at ICLR 2016. The announcement says its DDPG algorithm showed it was possible for deep reinforcement learning to venture into continuous control.

  8. As of October 2026

    Back at Google DeepMind

    He is a research scientist at Google DeepMind, his second stint at the lab after the years from 2014 to 2019.

Career

  1. Since 2014Google DeepMindResearch scientist
  2. From 2026Elefant AICo-founder and research team lead
  3. From 2019Twitter
  4. From 2013University of QueenslandResearcher

Education

  1. 2011University of QueenslandPhD in computational neuroscience
  2. 2006Massey UniversityBSc Hons in mathematical physics

From the biography.

  1. The median h-index is 48 at universities and 14 at frontier AI labsIncludes Jonathan J. Hunt. The median h-index is 48 at universities and 14 at frontier AI labs. We compare publishing careers, recent citations, dual appointments and job moves across 2,420 university researchers and 308 frontier-lab researchers.
  2. 11% of AI doctorate holders from 2010 to 2016 founded or led a company or lab by 10 years after graduationIncludes Jonathan J. Hunt. We estimate how often AI doctorate holders founded or led a company or lab by 10 years after graduation, how waiting times compare between founders and top jobs, and which universities trained the most.
  3. Google had the most people shaping AI every year from 2014 to 2025Includes Google. Our estimates of which organizations employed the most people shaping AI each year from 1955 to 2025, tracing the shift from university labs to companies.
  4. Google has the highest combined h-index among AI organizationsIncludes Google. We estimate that Google holds the highest combined h-index at 8,016, while universities account for 73% of the deduplicated total across AI organizations.

Colleagues

People named in this biography, and people with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Jonathan J. Hunt yet. Send us a free-licensed photo.

  1. jj hunt me.net.nz. Checked 21 October 2026.
  2. Sparse Coding Can Predict Primary Visual Cortex Receptive Field Changes Induced by Abnormal Visual Input ncbi.nlm.nih.gov. Checked 21 October 2026.
  3. Continuous control with deep reinforcement learning arxiv.org. Checked 21 October 2026.
  4. Successor Features for Transfer in Reinforcement Learning arxiv.org. Checked 21 October 2026.
  5. Hyperbolic Deep Reinforcement Learning arxiv.org. Checked 21 October 2026.
  6. ICLR Poster Hyperbolic Deep Reinforcement Learning iclr.cc. Checked 21 October 2026.
  7. Scaling Behavior Cloning Improves Causal Reasoning: An Open Model for Real-Time Video Game Playing arxiv.org. Checked 21 October 2026.
  8. Announcing the Test of Time Awards from ICLR 2016 blog.iclr.cc. Checked 21 October 2026.