Tom Schaul
Tom Schaul is a senior staff research scientist who leads the reinforcement learning team at Google DeepMind in London. He grew up in Luxembourg, studied computer science in Switzerland with exchanges at Waterloo and Columbia, and completed an MSc at EPFL in 2005. A PhD at TU Munich in 2011, under Jürgen Schmidhuber at the Swiss AI Lab IDSIA, was followed by a postdoc with Yann LeCun and a move to DeepMind in 2013. Prioritized Experience Replay appeared at ICLR in 2016, and Dueling Network Architectures for Deep Reinforcement Learning received an ICML best paper award in 2016. He coauthored the 2019 Nature paper in which AlphaStar was rated at Grandmaster level in StarCraft II.
- Now at
- Google DeepMind
- Works
- 140
- Citations
- 17,255
- h-index
- 40
Publication numbers from OpenAlex, as of 17 October 2026.
Timeline
-
2005
An EPFL master's degree
He grows up in Luxembourg and studies computer science in Switzerland, with exchanges at Waterloo and Columbia. He obtains an MSc from EPFL in 2005.
-
2011
A doctorate with Schmidhuber
He completes a PhD at TU Munich in 2011 under the supervision of Jürgen Schmidhuber at the Swiss AI Lab IDSIA. The thesis Studies in Continuous Black-box Optimization receives summa cum laude.
-
2011–2013
Postdoc with Yann LeCun
He works as a postdoc at the Courant Institute of NYU from 2011 to 2013, in the lab of Yann LeCun.
-
2013
Joining DeepMind
He joins DeepMind in 2013, while it is a start-up.
-
2016
Replay and a dueling network
With John Quan, Ioannis Antonoglou and David Silver, he publishes Prioritized Experience Replay at ICLR in 2016. The paper replays important transitions more often in Deep Q-Networks and beats uniform replay on 41 of 49 Atari games. Dueling Network Architectures for Deep Reinforcement Learning receives an ICML best paper award in 2016.
-
2019
Grandmaster play in StarCraft
He coauthors the 2019 Nature paper Grandmaster level in StarCraft II using multi-agent reinforcement learning. AlphaStar is rated at Grandmaster level for all three StarCraft races and above 99.8% of officially ranked human players.
-
2025–2026
Program chair for RLC
He serves as program chair for the 2026 Reinforcement Learning Conference. In December 2025 two of his papers appear at NeurIPS, on meta-learned training data curation and on the symmetry of plasticity and empowerment, with the empowerment paper a spotlight.
-
As of October 2026
Leading the RL team
He is a senior staff research scientist and leads the reinforcement learning team at Google DeepMind in London.
Sources and credits
Photo: We have no published photo of Tom Schaul yet. Send us a free-licensed photo.