Timezone: »
Successful coordination in Dec-POMDPs requires agents to adopt robust strategies and interpretable styles of play for their partner. A common failure mode is symmetry breaking, when agents arbitrarily converge on one out of many equivalent but mutually incompatible policies. Commonly these examples include partial observability, e.g. waving your right hand vs. left hand to convey a covert message. In this paper, we present a novel equivariant network architecture for use in Dec-POMDPs that prevents the agent from learning policies which break symmetries, doing so more effectively than prior methods. Our method also acts as a "coordination-improvement operator" for generic, pre-trained policies, and thus may be applied at test-time in conjunction with any self-play algorithm. We provide theoretical guarantees of our work and test on the AI benchmark task of Hanabi, where we demonstrate our methods outperforming other symmetry-aware baselines in zero-shot coordination, as well as able to improve the coordination ability of a variety of pre-trained policies. In particular, we show our method can be used to improve on the state of the art for zero-shot coordination on the Hanabi benchmark.
Author Information
Darius Muglich (University of Oxford)
Christian Schroeder de Witt (University of Oxford)
I am a 4th-year PhD student conducting fundamental algorithmic research in deep multi-agent reinforcement learning and climate change. My supervision is jointly between Prof. Shimon Whiteson (WhiRL - see my [profile](http://whirl.cs.ox.ac.uk/member/christian-schroeder-de-witt/)) and Prof. Philip Torr (Torr Vision Group).
Elise van der Pol (Microsoft Research)
Shimon Whiteson (Oxford University)
Jakob Foerster (University of Oxford)
Jakob Foerster received a CIFAR AI chair in 2019 and is starting as an Assistant Professor at the University of Toronto and the Vector Institute in the academic year 20/21. During his PhD at the University of Oxford, he helped bring deep multi-agent reinforcement learning to the forefront of AI research and interned at Google Brain, OpenAI, and DeepMind. He has since been working as a research scientist at Facebook AI Research in California, where he will continue advancing the field up to his move to Toronto. He was the lead organizer of the first Emergent Communication (EmeCom) workshop at NeurIPS in 2017, which he has helped organize ever since.
More from the Same Authors
-
2021 : Grounding Aleatoric Uncertainty in Unsupervised Environment Design »
Minqi Jiang · Michael Dennis · Jack Parker-Holder · Andrei Lupu · Heinrich Kuttler · Edward Grefenstette · Tim Rocktäschel · Jakob Foerster -
2021 : No DICE: An Investigation of the Bias-Variance Tradeoff in Meta-Gradients »
Risto Vuorio · Jacob Beck · Greg Farquhar · Jakob Foerster · Shimon Whiteson -
2021 : That Escalated Quickly: Compounding Complexity by Editing Levels at the Frontier of Agent Capabilities »
Jack Parker-Holder · Minqi Jiang · Michael Dennis · Mikayel Samvelyan · Jakob Foerster · Edward Grefenstette · Tim Rocktäschel -
2021 : A Fine-Tuning Approach to Belief State Modeling »
Samuel Sokota · Hengyuan Hu · David Wu · Jakob Foerster · Noam Brown -
2021 : Generalized Belief Learning in Multi-Agent Settings »
Darius Muglich · Luisa Zintgraf · Christian Schroeder de Witt · Shimon Whiteson · Jakob Foerster -
2022 : Adversarial Cheap Talk »
Chris Lu · Timon Willi · Alistair Letcher · Jakob Foerster -
2022 : Human-AI Coordination via Human-Regularized Search and Learning »
Hengyuan Hu · David Wu · Adam Lerer · Jakob Foerster · Noam Brown -
2022 : Adversarial Cheap Talk »
Chris Lu · Timon Willi · Alistair Letcher · Jakob Foerster -
2022 : MAESTRO: Open-Ended Environment Design for Multi-Agent Reinforcement Learning »
Mikayel Samvelyan · Akbir Khan · Michael Dennis · Minqi Jiang · Jack Parker-Holder · Jakob Foerster · Roberta Raileanu · Tim Rocktäschel -
2023 Poster: Recurrent Hypernetworks are Surprisingly SOTA in Meta-RL »
Jacob Beck · Risto Vuorio · Zheng Xiong · Shimon Whiteson -
2023 Poster: Discovering General Reinforcement Learning Algorithms with Adversarial Environment Design »
Matthew T Jackson · Minqi Jiang · Jack Parker-Holder · Risto Vuorio · Chris Lu · Greg Farquhar · Shimon Whiteson · Jakob Foerster -
2023 Poster: Similarity-based cooperative equilibrium »
Caspar Oesterheld · Johannes Treutlein · Roger Grosse · Vincent Conitzer · Jakob Foerster -
2023 Poster: Structured State Space Models for In-Context Reinforcement Learning »
Chris Lu · Yannick Schroecker · Albert Gu · Emilio Parisotto · Jakob Foerster · Satinder Singh · Feryal Behbahani -
2023 Poster: SMACv2: An Improved Benchmark for Cooperative Multi-Agent Reinforcement Learning »
Benjamin Ellis · Jonathan Cook · Skander Moalla · Mikayel Samvelyan · Mingfei Sun · Anuj Mahajan · Jakob Foerster · Shimon Whiteson -
2023 Poster: The Waymo Open Sim Agents Challenge »
Nico Montali · John Lambert · Paul Mougin · Alex Kuefler · Nicholas Rhinehart · Michelle Li · Cole Gulino · Tristan Emrich · Zoey Yang · Shimon Whiteson · Brandyn White · Dragomir Anguelov -
2023 Workshop: Multi-Agent Security »
Christian Schroeder de Witt · Hawra Milani · Klaudia Krawiecka · Swapneel Mehta · Carla Cremer · Martin Strohmeier -
2023 Workshop: Socially Responsible Language Modelling Research (SoLaR) »
Usman Anwar · David Krueger · Samuel Bowman · Jakob Foerster · Su Lin Blodgett · Roberta Raileanu · Alan Chan · Katherine Lee · Laura Ruis · Robert Kirk · Yawen Duan · Xin Chen · Kawin Ethayarajh -
2022 : Jakob Foerster »
Jakob Foerster -
2022 Poster: Proximal Learning With Opponent-Learning Awareness »
Stephen Zhao · Chris Lu · Roger Grosse · Jakob Foerster -
2022 Poster: Nocturne: a scalable driving benchmark for bringing multi-agent learning one step closer to the real world »
Eugene Vinitsky · Nathan Lichtlé · Xiaomeng Yang · Brandon Amos · Jakob Foerster -
2022 Poster: In Defense of the Unitary Scalarization for Deep Multi-Task Learning »
Vitaly Kurin · Alessandro De Palma · Ilya Kostrikov · Shimon Whiteson · Pawan K Mudigonda -
2022 Poster: Grounding Aleatoric Uncertainty for Unsupervised Environment Design »
Minqi Jiang · Michael Dennis · Jack Parker-Holder · Andrei Lupu · Heinrich Küttler · Edward Grefenstette · Tim Rocktäschel · Jakob Foerster -
2022 Poster: Off-Team Learning »
Brandon Cui · Hengyuan Hu · Andrei Lupu · Samuel Sokota · Jakob Foerster -
2022 Poster: Self-Explaining Deviations for Coordination »
Hengyuan Hu · Samuel Sokota · David Wu · Anton Bakhtin · Andrei Lupu · Brandon Cui · Jakob Foerster -
2022 Poster: Discovered Policy Optimisation »
Chris Lu · Jakub Kuba · Alistair Letcher · Luke Metz · Christian Schroeder de Witt · Jakob Foerster -
2022 Poster: Maximum Class Separation as Inductive Bias in One Matrix »
Tejaswi Kasarla · Gertjan Burghouts · Max van Spengler · Elise van der Pol · Rita Cucchiara · Pascal Mettes -
2022 Poster: Influencing Long-Term Behavior in Multiagent Reinforcement Learning »
Dong-Ki Kim · Matthew Riemer · Miao Liu · Jakob Foerster · Michael Everett · Chuangchuang Sun · Gerald Tesauro · Jonathan How -
2021 Workshop: Cooperative AI »
Natasha Jaques · Edward Hughes · Jakob Foerster · Noam Brown · Kalesha Bullard · Charlotte Smith -
2021 Workshop: Ecological Theory of Reinforcement Learning: How Does Task Design Influence Agent Learning? »
Manfred Díaz · Hiroki Furuta · Elise van der Pol · Lisa Lee · Shixiang (Shane) Gu · Pablo Samuel Castro · Simon Du · Marc Bellemare · Sergey Levine -
2021 Poster: Replay-Guided Adversarial Environment Design »
Minqi Jiang · Michael Dennis · Jack Parker-Holder · Jakob Foerster · Edward Grefenstette · Tim Rocktäschel -
2021 Poster: K-level Reasoning for Zero-Shot Coordination in Hanabi »
Brandon Cui · Hengyuan Hu · Luis Pineda · Jakob Foerster -
2021 Poster: Neural Pseudo-Label Optimism for the Bank Loan Problem »
Aldo Pacchiano · Shaun Singh · Edward Chou · Alex Berg · Jakob Foerster -
2020 Workshop: Talking to Strangers: Zero-Shot Emergent Communication »
Marie Ossenkopf · Angelos Filos · Abhinav Gupta · Michael Noukhovitch · Angeliki Lazaridou · Jakob Foerster · Kalesha Bullard · Rahma Chaabouni · Eugene Kharitonov · Roberto Dessì -
2020 Poster: Ridge Rider: Finding Diverse Solutions by Following Eigenvectors of the Hessian »
Jack Parker-Holder · Luke Metz · Cinjon Resnick · Hengyuan Hu · Adam Lerer · Alistair Letcher · Alexander Peysakhovich · Aldo Pacchiano · Jakob Foerster -
2020 Poster: MDP Homomorphic Networks: Group Symmetries in Reinforcement Learning »
Elise van der Pol · Daniel E Worrall · Herke van Hoof · Frans Oliehoek · Max Welling -
2019 : Bayes-Adaptive Deep Reinforcement Learning via Meta-Learning - Invited Talk »
Shimon Whiteson -
2019 Workshop: Emergent Communication: Towards Natural Language »
Abhinav Gupta · Michael Noukhovitch · Cinjon Resnick · Natasha Jaques · Angelos Filos · Marie Ossenkopf · Angeliki Lazaridou · Jakob Foerster · Ryan Lowe · Douwe Kiela · Kyunghyun Cho -
2019 Poster: Loaded DiCE: Trading off Bias and Variance in Any-Order Score Function Gradient Estimators for Reinforcement Learning »
Gregory Farquhar · Shimon Whiteson · Jakob Foerster -
2019 Poster: Multi-Agent Common Knowledge Reinforcement Learning »
Christian Schroeder de Witt · Jakob Foerster · Gregory Farquhar · Philip Torr · Wendelin Boehmer · Shimon Whiteson -
2019 Poster: Hyperspherical Prototype Networks »
Pascal Mettes · Elise van der Pol · Cees Snoek -
2018 Workshop: Emergent Communication Workshop »
Jakob Foerster · Angeliki Lazaridou · Ryan Lowe · Igor Mordatch · Douwe Kiela · Kyunghyun Cho -
2017 Workshop: Emergent Communication Workshop »
Jakob Foerster · Igor Mordatch · Angeliki Lazaridou · Kyunghyun Cho · Douwe Kiela · Pieter Abbeel -
2017 Poster: Dynamic-Depth Context Tree Weighting »
Joao V Messias · Shimon Whiteson -
2016 Poster: Learning to Communicate with Deep Multi-Agent Reinforcement Learning »
Jakob Foerster · Yannis Assael · Nando de Freitas · Shimon Whiteson