Pith. sign in

REVIEW 2 cited by

MEAL: A Benchmark for Continual Multi-Agent Reinforcement Learning

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2506.14990 v3 pith:6OYASPGP submitted 2025-06-17 cs.AI

classification cs.AI
keywords learningcontinualmulti-agentmealsequencesbenchmarkenvironmentsreinforcement
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Benchmarks play a central role in reinforcement learning (RL) research, yet their computational constraints often shape what is studied. Despite the motivation of lifelong learning, most continual RL papers consider only 3-10 sequential tasks, as CPU-bound environments make longer sequences impractical. Meanwhile, continual learning in cooperative multi-agent settings remains largely unexplored. To address these gaps, we introduce MEAL (Multi-agent Environments for Adaptive Learning), the first benchmark for continual multi-agent RL. By leveraging JAX and GPU acceleration, MEAL enables training on sequences of 100 tasks in a few hours on a single GPU. We find that long task sequences reveal failure modes that do not appear at smaller scales.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Offline Multi-agent Continual Cooperation via Skill Partition and Reuse

    cs.AI 2026-06 unverdicted novelty 7.0 of 10

    COMAD discovers and reuses coordination skills from mixed offline MARL data via auto-encoders and density-based estimation to achieve continual learning with better transfer.

  2. Stagnant Neuron: Towards Understanding the Plasticity Loss in Multi-Agent Reinforcement Learning Value Factorization Methods

    cs.LG 2026-06 unverdicted novelty 6.0 of 10

    KNIFE targets stagnant neurons in MARL value factorization by replacing them with a composite of frozen, re-initialized, and compensating units to restore plasticity while preserving cooperation knowledge.

Reference graph

Works this paper leans on

51 extracted references · 24 canonical work pages · cited by 2 Pith papers

  1. [1]

    Loss of plasticity in continual deep reinforcement learning

    Zaheer Abbas, Rosie Zhao, Joseph Modayil, Adam White, and Marlos C Machado. Loss of plasticity in continual deep reinforcement learning. InConference on lifelong learning agents, pages 620–636. PMLR, 2023

  2. [2]

    Melting pot 2.0.arXiv preprint arXiv:2211.13746, 2022

    John P Agapiou, Alexander Sasha Vezhnevets, Edgar A Duéñez-Guzmán, Jayd Matyas, Yi- ran Mao, Peter Sunehag, Raphael Köster, Udari Madhushani, Kavya Kopparapu, Ramona Comanescu, et al. Melting pot 2.0.arXiv preprint arXiv:2211.13746, 2022

  3. [3]

    Memory aware synapses: Learning what (not) to forget

    Rahaf Aljundi, Francesca Babiloni, Mohamed Elhoseiny, Marcus Rohrbach, and Tinne Tuyte- laars. Memory aware synapses: Learning what (not) to forget. InProceedings of the European conference on computer vision (ECCV), pages 139–154, 2018

  4. [4]

    Pseudo-rehearsal: Achieving deep reinforcement learning without catastrophic forgetting.Neurocomputing, 428: 291–307, 2021

    Craig Atkinson, Brendan McCane, Lech Szymanski, and Anthony Robins. Pseudo-rehearsal: Achieving deep reinforcement learning without catastrophic forgetting.Neurocomputing, 428: 291–307, 2021

  5. [5]

    Curriculum learning

    Yoshua Bengio, Jérôme Louradour, Ronan Collobert, and Jason Weston. Curriculum learning. InProceedings of the 26th annual international conference on machine learning, pages 41–48, 2009

  6. [6]

    JAX: composable transformations of Python+NumPy programs, 2018

    James Bradbury, Roy Frostig, Peter Hawkins, Matthew James Johnson, Chris Leary, Dougal Maclaurin, George Necula, Adam Paszke, Jake VanderPlas, Skye Wanderman-Milne, and Qiao Zhang. JAX: composable transformations of Python+NumPy programs, 2018. URL http://github.com/jax-ml/jax

  7. [7]

    Ho, Thomas Griffiths, Sanjit Seshia, Pieter Abbeel, and Anca Dragan

    Micah Carroll, Rohin Shah, Mark K. Ho, Thomas Griffiths, Sanjit Seshia, Pieter Abbeel, and Anca Dragan. On the utility of learning about humans for human-ai coordination. InAdvances in Neural Information Processing Systems (NeurIPS), vol- ume 32, 2019. URL https://proceedings.neurips.cc/paper_files/paper/2019/ file/f5b1b89d3db40d65b49f8f9e383ac5dd-Paper.pdf

  8. [8]

    Efficient lifelong learning with a-gem.arXiv preprint arXiv:1812.00420, 2018

    Arslan Chaudhry, Marc’Aurelio Ranzato, Marcus Rohrbach, and Mohamed Elhoseiny. Efficient lifelong learning with a-gem.arXiv preprint arXiv:1812.00420, 2018

Show all 51 references
  1. [9]

    Stable continual reinforcement learning via diffusion-based trajectory replay.arXiv preprint arXiv:2411.10809, 2024

    Feng Chen, Fuguang Han, Cong Guan, Lei Yuan, Zhilong Zhang, Yang Yu, and Zongzhang Zhang. Stable continual reinforcement learning via diffusion-based trajectory replay.arXiv preprint arXiv:2411.10809, 2024

  2. [10]

    Parseval regularization for continual reinforcement learning.Advances in Neural Information Processing Systems, 37: 127937–127967, 2024

    Wesley Chung, Lynn Cherif, Doina Precup, and David Meger. Parseval regularization for continual reinforcement learning.Advances in Neural Information Processing Systems, 37: 127937–127967, 2024

  3. [11]

    Is independent learning all you need in the starcraft multi-agent challenge?arXiv preprint arXiv:2011.09533, 2020

    Christian Schroeder De Witt, Tarun Gupta, Denys Makoviichuk, Viktor Makoviychuk, Philip HS Torr, Mingfei Sun, and Shimon Whiteson. Is independent learning all you need in the starcraft multi-agent challenge?arXiv preprint arXiv:2011.09533, 2020

  4. [12]

    Loss of plasticity in deep continual learning.Nature, 632 (8026):768–774, 2024

    Shibhansh Dohare, J Fernando Hernandez-Garcia, Qingfeng Lan, Parash Rahman, A Rupam Mahmood, and Richard S Sutton. Loss of plasticity in deep continual learning.Nature, 632 (8026):768–774, 2024

  5. [13]

    Smacv2: An improved benchmark for cooperative multi-agent reinforcement learning.Advances in Neural Information Processing Systems, 36:37567–37593, 2023

    Benjamin Ellis, Jonathan Cook, Skander Moalla, Mikayel Samvelyan, Mingfei Sun, Anuj Mahajan, Jakob Foerster, and Shimon Whiteson. Smacv2: An improved benchmark for cooperative multi-agent reinforcement learning.Advances in Neural Information Processing Systems, 36:37567–37593,...

  6. [14]

    Continual reinforcement learning via autoencoder-driven task and new environment recognition

    Zeki Doruk Erden, Donia Gasmi, and Boi Faltings. Continual reinforcement learning via autoencoder-driven task and new environment recognition. InThe Seventeenth Workshop on Adaptive and Learning Agents

  7. [15]

    Counterfactual multi-agent policy gradients

    Jakob Foerster, Gregory Farquhar, Triantafyllos Afouras, Nantas Nardelli, and Shimon Whiteson. Counterfactual multi-agent policy gradients. InProceedings of the AAAI conference on artificial intelligence, volume 32, 2018

  8. [16]

    Learning with opponent-learning awareness.arXiv preprint arXiv:1709.04326, 2017

    Jakob N Foerster, Richard Y Chen, Maruan Al-Shedivat, Shimon Whiteson, Pieter Abbeel, and Igor Mordatch. Learning with opponent-learning awareness.arXiv preprint arXiv:1709.04326, 2017

  9. [17]

    Continual deep reinforcement learning with task-agnostic policy distillation.Scientific Reports, 14(1):31661, 2024

    Muhammad Burhan Hafez and Kerim Erekmen. Continual deep reinforcement learning with task-agnostic policy distillation.Scientific Reports, 14(1):31661, 2024

  10. [18]

    A survey and critique of multiagent deep reinforcement learning.Autonomous Agents and Multi-Agent Systems, 33(6):750–797, 2019

    Pablo Hernandez-Leal, Bilal Kartal, and Matthew E Taylor. A survey and critique of multiagent deep reinforcement learning.Autonomous Agents and Multi-Agent Systems, 33(6):750–797, 2019

  11. [19]

    other-play

    Hengyuan Hu, Adam Lerer, Alex Peysakhovich, and Jakob Foerster. “other-play” for zero-shot coordination. InInternational Conference on Machine Learning, pages 4399–4410. PMLR, 2020

  12. [20]

    Inequity aversion improves cooperation in intertemporal social dilemmas.Advances in neural information processing systems, 31, 2018

    Edward Hughes, Joel Z Leibo, Matthew Phillips, Karl Tuyls, Edgar Dueñez-Guzman, Antonio García Castañeda, Iain Dunning, Tina Zhu, Kevin McKee, Raphael Koster, et al. Inequity aversion improves cooperation in intertemporal social dilemmas.Advances in neural information processi...

  13. [21]

    Overcoming catastrophic forgetting in neural networks.Proceedings of the national academy of sciences, 114(13):3521–3526, 2017

    James Kirkpatrick, Razvan Pascanu, Neil Rabinowitz, Joel Veness, Guillaume Desjardins, Andrei A Rusu, Kieran Milan, John Quan, Tiago Ramalho, Agnieszka Grabska-Barwinska, et al. Overcoming catastrophic forgetting in neural networks.Proceedings of the national academy of scienc...

  14. [22]

    Evaluating the robustness of collaborative agents.arXiv preprint arXiv:2101.05507, 2021

    Paul Knott, Micah Carroll, Sam Devlin, Kamil Ciosek, Katja Hofmann, Anca D Dragan, and Rohin Shah. Evaluating the robustness of collaborative agents.arXiv preprint arXiv:2101.05507, 2021

  15. [23]

    Google research football: A novel reinforcement learning environment

    Karol Kurach, Anton Raichuk, Piotr Sta´nczyk, Michał Zaj ˛ ac, Olivier Bachem, Lasse Espeholt, Carlos Riquelme, Damien Vincent, Marcin Michalski, Olivier Bousquet, et al. Google research football: A novel reinforcement learning environment. InProceedings of the AAAI conference...

  16. [24]

    Multi-agent actor-critic for mixed cooperative-competitive environments.Advances in neural information processing systems, 30, 2017

    Ryan Lowe, Yi I Wu, Aviv Tamar, Jean Harb, OpenAI Pieter Abbeel, and Igor Mordatch. Multi-agent actor-critic for mixed cooperative-competitive environments.Advances in neural information processing systems, 30, 2017

  17. [25]

    Packnet: Adding multiple tasks to a single network by iterative pruning

    Arun Mallya and Svetlana Lazebnik. Packnet: Adding multiple tasks to a single network by iterative pruning. InProceedings of the IEEE conference on Computer Vision and Pattern Recognition, pages 7765–7773, 2018

  18. [26]

    Flatland-rl: Multi-agent reinforcement learning on trains.arXiv preprint arXiv:2012.05893, 2020

    Sharada Mohanty, Erik Nygren, Florian Laurent, Manuel Schneider, Christian Scheller, Nilabha Bhattacharya, Jeremy Watson, Adrian Egli, Christian Eichenberger, Christian Baumberger, et al. Flatland-rl: Multi-agent reinforcement learning on trains.arXiv preprint arXiv:2012.05893, 2020

  19. [27]

    Emergence of grounded compositional language in multi- agent populations

    Igor Mordatch and Pieter Abbeel. Emergence of grounded compositional language in multi- agent populations. InProceedings of the AAAI conference on artificial intelligence, volume 32, 2018

  20. [28]

    Curriculum learning for reinforcement learning domains: A framework and survey.Journal of Machine Learning Research, 21(181):1–50, 2020

    Sanmit Narvekar, Bei Peng, Matteo Leonetti, Jivko Sinapov, Matthew E Taylor, and Peter Stone. Curriculum learning for reinforcement learning domains: A framework and survey.Journal of Machine Learning Research, 21(181):1–50, 2020. 12

  21. [29]

    A review of cooperative multi-agent deep reinforcement learning.arXiv preprint arXiv:1908.03963, 2019

    Afshin OroojlooyJadid and Davood Hajinezhad. A review of cooperative multi-agent deep reinforcement learning.arXiv preprint arXiv:1908.03963, 2019

  22. [30]

    A multi-agent reinforcement learning model of common-pool resource appropriation.Advances in neural information processing systems, 30, 2017

    Julien Perolat, Joel Z Leibo, Vinicius Zambaldi, Charles Beattie, Karl Tuyls, and Thore Graepel. A multi-agent reinforcement learning model of common-pool resource appropriation.Advances in neural information processing systems, 30, 2017

  23. [31]

    Teacher algorithms for curriculum learning of deep rl in continuously parameterized environments

    Rémy Portelas, Cédric Colas, Katja Hofmann, and Pierre-Yves Oudeyer. Teacher algorithms for curriculum learning of deep rl in continuously parameterized environments. InConference on Robot Learning, pages 835–853. PMLR, 2020

  24. [32]

    Cora: Benchmarks, baselines, and metrics as a platform for continual reinforcement learning agents

    Sam Powers, Eliot Xing, Eric Kolve, Roozbeh Mottaghi, and Abhinav Gupta. Cora: Benchmarks, baselines, and metrics as a platform for continual reinforcement learning agents. InConference on Lifelong Learning Agents, pages 705–743. PMLR, 2022

  25. [33]

    Monotonic value function factorisation for deep multi-agent reinforcement learning.Journal of Machine Learning Research, 21(178):1–51, 2020

    Tabish Rashid, Mikayel Samvelyan, Christian Schroeder De Witt, Gregory Farquhar, Jakob Foerster, and Shimon Whiteson. Monotonic value function factorisation for deep multi-agent reinforcement learning.Journal of Machine Learning Research, 21(178):1–51, 2020

  26. [34]

    Pommerman: A multi-agent playground.arXiv preprint arXiv:1809.07124, 2018

    Cinjon Resnick, Wes Eldridge, David Ha, Denny Britz, Jakob Foerster, Julian Togelius, Kyunghyun Cho, and Joan Bruna. Pommerman: A multi-agent playground.arXiv preprint arXiv:1809.07124, 2018

  27. [35]

    Jaxmarl: Multi-agent rl environments and algorithms in jax

    Alexander Rutherford, Benjamin Ellis, Matteo Gallici, Jonathan Cook, Andrei Lupu, Garðar Ingvarsson, Timon Willi, Akbir Khan, Christian Schroeder de Witt, Alexandra Souly, et al. Jaxmarl: Multi-agent rl environments and algorithms in jax. InProceedings of the 23rd Internationa...

  28. [36]

    The starcraft multi-agent challenge.arXiv preprint arXiv:1902.04043, 2019

    Mikayel Samvelyan, Tabish Rashid, Christian Schroeder De Witt, Gregory Farquhar, Nan- tas Nardelli, Tim GJ Rudner, Chia-Man Hung, Philip HS Torr, Jakob Foerster, and Shimon Whiteson. The starcraft multi-agent challenge.arXiv preprint arXiv:1902.04043, 2019

  29. [37]

    Continual and multi-task reinforcement learning with shared episodic memory.arXiv preprint arXiv:1905.02662, 2019

    Artyom Y Sorokin and Mikhail S Burtsev. Continual and multi-task reinforcement learning with shared episodic memory.arXiv preprint arXiv:1905.02662, 2019

  30. [38]

    Collaborating with humans without human data.Advances in Neural Information Processing Systems, 34: 14502–14515, 2021

    DJ Strouse, Kevin McKee, Matt Botvinick, Edward Hughes, and Richard Everett. Collaborating with humans without human data.Advances in Neural Information Processing Systems, 34: 14502–14515, 2021

  31. [39]

    Value- decomposition networks for cooperative multi-agent learning.arXiv preprint arXiv:1706.05296, 2017

    Peter Sunehag, Guy Lever, Audrunas Gruslys, Wojciech Marian Czarnecki, Vinicius Zambaldi, Max Jaderberg, Marc Lanctot, Nicolas Sonnerat, Joel Z Leibo, Karl Tuyls, et al. Value- decomposition networks for cooperative multi-agent learning.arXiv preprint arXiv:1706.05296, 2017

  32. [40]

    Coom: a game benchmark for continual reinforcement learning.Advances in Neural Information Processing Systems, 36, 2023

    Tristan Tomilin, Meng Fang, Yudi Zhang, and Mykola Pechenizkiy. Coom: a game benchmark for continual reinforcement learning.Advances in Neural Information Processing Systems, 36, 2023

  33. [41]

    Contin- ual world: A robotic benchmark for continual reinforcement learning.Advances in Neural Information Processing Systems, 34:28496–28510, 2021

    Maciej Wołczyk, Michał Zaj ˛ ac, Razvan Pascanu, Łukasz Kuci´nski, and Piotr Miło´s. Contin- ual world: A robotic benchmark for continual reinforcement learning.Advances in Neural Information Processing Systems, 34:28496–28510, 2021

  34. [42]

    Too many cooks: Bayesian inference for coordinating multi-agent collaboration.Topics in Cognitive Science, 13(2):414–432, 2021

    Sarah A Wu, Rose E Wang, James A Evans, Joshua B Tenenbaum, David C Parkes, and Max Kleiman-Weiner. Too many cooks: Bayesian inference for coordinating multi-agent collaboration.Topics in Cognitive Science, 13(2):414–432, 2021

  35. [43]

    The surprising effectiveness of ppo in cooperative multi-agent games.Advances in neural information processing systems, 35:24611–24624, 2022

    Chao Yu, Akash Velu, Eugene Vinitsky, Jiaxuan Gao, Yu Wang, Alexandre Bayen, and Yi Wu. The surprising effectiveness of ppo in cooperative multi-agent games.Advances in neural information processing systems, 35:24611–24624, 2022

  36. [44]

    A survey of progress on cooperative multi-agent reinforcement learning in open environment.arXiv preprint arXiv:2312.01058, 2023

    Lei Yuan, Ziqian Zhang, Lihe Li, Cong Guan, and Yang Yu. A survey of progress on cooperative multi-agent reinforcement learning in open environment.arXiv preprint arXiv:2312.01058, 2023. 13

  37. [45]

    Multiagent continual coordination via progressive task contextualization.IEEE Transactions on Neural Networks and Learning Systems, 2024

    Lei Yuan, Lihe Li, Ziqian Zhang, Fuxiang Zhang, Cong Guan, and Yang Yu. Multiagent continual coordination via progressive task contextualization.IEEE Transactions on Neural Networks and Learning Systems, 2024

  38. [46]

    Continual learning through synaptic intelligence

    Friedemann Zenke, Ben Poole, and Surya Ganguli. Continual learning through synaptic intelligence. InInternational conference on machine learning, pages 3987–3995. PMLR, 2017. 14 A Implementation Details A.1 Maximum Soup Delivery Calculator Let a kitchen layout L be defined by ...

  39. [47]

    Draw size.Sample h, wand create an h×w matrix initialised toFLOORtiles, then overwrite the border withWALL

  40. [48]

    Place interactive tiles.For each symbol in {GOAL,POT,ONION_PILE,PLATE_PILE} choose a random multiplicity m∈ {1,2}and stamp the symbol onto m uniformly chosen floor cells

  41. [49]

    Place nadd additional walls on random floor cells

    Inject extra walls.Let ntarget = ρ Nint and nadd = max 0, ntarget −N unpass(G) . Place nadd additional walls on random floor cells. 4.Place agents.Stampn a AGENTsymbols on random remaining floor cells

  42. [50]

    Validate.Run the deterministic evaluate_grid solver; if it returns True, terminate and return(G), otherwise restart

  43. [51]

    7.Return.Output the final grid

    Cleanup.Remove any interactive elements and tiles that are unreachable from all agent positions. 7.Return.Output the final grid. Solvability criterion.The validator (Appendix A.3) checks (i) path connectivity between every agent and each interactive tile family, (ii) at least ...

Pith tools