Pith. sign in

REVIEW 3 major objections 6 minor 1 cited by

R2BC: Multi-Agent Imitation Learning from Single-Agent Demonstrations

T0 review · 3 major / 6 minor · reviewed 2026-08-04 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Round-Robin Behavior Cloning (R2BC) lets one human teach a robot team by teleoperating one agent at a time, matching or beating behavior cloning from privileged synchronized demonstrations.

desk verdict R2BC is a genuinely new problem setup with solid empirical support, but the paper overstates one real-world result and needs a softer conclusion. read the letter →

arxiv 2510.18085 v2 pith:6SNHEQ6Z submitted 2025-10-20 cs.RO cs.AIcs.MA

Connor Mattson , Varun Raveendra , Ellen Novoseller , Nicholas Waytowich , Vernon J. Lawhern , Daniel S. Brown This is my paper · ORCID
classification cs.ROcs.AIcs.MA
keywords multi-agentimitationlearningbehaviorcloningsingle-agentdemonstrationsround-robintrainingmulti-robotteleoperationcovariateshiftcooperativesystemssim-to-realtransfer
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Multi-agent imitation learning has usually assumed that a demonstrator can provide synchronized, near-optimal actions for every agent at once—something a single human cannot actually do. This paper introduces Round-Robin Behavior Cloning (R2BC), in which a human teleoperates one agent at a time while the other agents execute their current learned policies, cycling through the team. The paper shows that this online loop matches or exceeds the performance of an oracle behavior cloning method trained on privileged synchronized joint demonstrations across four simulated cooperative tasks, and that policies trained from real human demonstrations outperform centralized behavior cloning by 3.25x and 5.9x on two physical robot tasks. The authors argue the reason is that the imperfect actions of non-demonstrated agents inject state diversity into the collected data, reducing covariate shift in the same way explicit noise injection does in single-agent methods. The result matters because it replaces an infeasible demonstration protocol with one any single operator can execute.

What carries the argument

Round-Robin Behavior Cloning (R2BC), a training loop that cycles which single agent the demonstrator controls while the rest of the team executes its current policies, then clones each agent's demonstrations into a separate policy trained only on that agent's buffer. The load-bearing mechanism is teammate-induced state perturbation: because the non-demonstrated agents act suboptimally and are themselves being updated, the demonstrated agent sees a diverse stream of teammate behavior, which teaches corrective actions and reduces covariate shift the way explicit noise injection does in single-agent online imitation learning.

What would settle it

Track the per-agent train-test loss gap and the change in teammate policies across successive rounds on a fixed task. If the loss gap grows or oscillates without bound as demonstrations accumulate—or if the gap does not shrink relative to an offline variant—the online co-adaptation loop is not converging and R2BC's advantage over joint behavior cloning would not be explained by covariate-shift reduction.

Watch

Extended reading notes

Core claim

R2BC's central claim is that a single human can train a coherent multi-agent policy through sequential single-agent demonstrations, provided the demonstrations are collected online in a round-robin loop: agent i is teleoperated while the other agents run their current cloned policies; the resulting observation-action pairs go into agent i's own buffer; and all policies are re-trained at regular intervals. Because the demonstrating agent is repeatedly exposed to states created by imperfect teammates, it collects corrective examples from a distribution close to what the team will actually experience at execution time. The paper argues this online co-adaptation, together with full decentralizat

Load-bearing premise

The round-robin online loop is assumed to converge: as each agent's policy is updated, the non-demonstrated agents become a moving target, and demonstrations collected against them still support a coherent team policy—but the paper provides no convergence or stability guarantee.

Editorial extensions

If this is right

  • A single human operator can train a multi-robot team with ordinary single-agent teleoperation; synchronized joint-action demonstrations are no longer a prerequisite for multi-agent imitation learning.
  • R2BC matches or surpasses behavior cloning from privileged joint demonstrations on four simulated cooperative tasks, so realistic demo collection does not have to cost task performance.
  • Policies trained from real human demonstrations and deployed on physical robots outperform centralized behavior cloning by 3.25x on navigation and 5.9x on block pushing, and short human interventions amplify the gap.
  • Ablations show that both decentralization and the online update loop are necessary—centralizing the policy or freezing teammate policies during collection degrades results—so the claimed mechanism is the online co-adaptation, not simply round-robin data collection.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The mechanism suggests a broader principle: a learning team can serve as its own data generator, with teammates' imperfections providing the corrective signal a teacher needs—so the teacher's update rate only needs to outpace teammate deterioration, a rate condition the paper does not analyze.
  • A testable extension would sweep the update frequency and round-robin order; if teammate-induced state diversity is the source of the benefit, the method should be robust to both, and the optimal schedule could guide deployment on larger teams.
  • The physical-robot demonstrations were collected by the paper's authors; an independent user study with non-expert teachers would test whether the 3.25x and 5.9x margins persist—a step the paper lists as future work.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The paper introduces Round-Robin Behavior Cloning (R2BC), a multi-agent imitation learning method that trains a team of agents from single-agent demonstrations provided sequentially by one human. While one agent is teleoperated, the other agents act according to their current learned policies; demonstrations are stored in per-agent buffers and policies are updated periodically in a round-robin fashion. The method is evaluated on four VMAS simulated tasks (Navigation, Balance, Buzz Wire, Transport) against three oracle baselines: Joint Behavior Cloning (JBC), DAgger, and DART, all of which use privileged centralized demonstrations from a MAPPO expert. R2BC matches or exceeds these baselines. Ablations show that both decentralization and online learning are essential. The method is also deployed on physical HeRo+ robots in navigation and block-pushing tasks with real human demonstrations, where R2BC is reported to outperform centralized JBC by factors of 3.25x and 5.9x, respectively. The central claim is that R2BC provides a practical alternative to coordinated joint-action demonstrations for multi-agent IL.

Significance. The problem setting—multi-agent IL from single-agent demonstrations—is novel, well-motivated, and practically relevant. The paper provides a simple, effective algorithm and backs it with a solid empirical methodology: 10 seeds with 99% confidence intervals in simulation, multiple ablations, physical robot experiments with human demonstrations, and a public supplement with code and videos. The result that a single human can train a cooperative team without synchronous joint-action demonstrations is an important step toward real-world multi-robot teaching. The main weakness is the absence of a convergence or stability analysis for the online co-adaptation loop, which is explicitly acknowledged in the conclusion. The empirical evidence is strong within the tested finite demonstration budgets, but the generality of the method beyond these budgets remains unexamined. Overall, the paper is a meaningful contribution to multi-agent IL and human-robot interaction, provided the authors address the concerns below.

major comments (3)
  1. [Algorithm 1, Section IV, and Section VIII] The core mechanism is an online alternating-update process: each agent's demonstrations are collected while teammates execute their current policies, and policies are updated after every k rounds. This is a moving-target problem, and the paper provides no analysis of whether the process converges, cycles, or diverges if run beyond the fixed demonstration budgets. The conclusion explicitly concedes the lack of theoretical guarantees. For the central claim that R2BC 'enables a single human operator to effectively train multi-robot systems,' this is a load-bearing gap: the reported success could, in principle, be an artifact of early stopping before the co-adaptation destabilizes, especially in coupled-dynamics tasks like Buzz Wire. I recommend either (a) adding experiments that run substantially longer than the current budgets (e.g., 2–4x more demonstrations) and report whether performance
  2. [Section V-C, Figure 2] The comparison to DAgger and DART, which the text uses to support the claim that R2BC achieves 'a similar reduction in compounding errors,' is not precisely specified. It is stated that these baselines use 'oracle centralized corrections generated from the same multi-agent RL policy,' but it is unclear whether (i) the policies are decentralized or centralized, (ii) the expert provides joint-action labels at every timestep for all agents, and (iii) the same demonstration budget and round-robin schedule are used. If the baselines receive privileged joint-action corrections at every step, they are strong oracles, but this should be stated explicitly. If they instead use single-agent corrections, the setup should be described. Without this detail, the reader cannot assess whether the parity result is due to R2BC's specific mechanism or to the online correction paradigm generally. Please prov
  3. [Abstract and Section VII-B, Figure 6] The abstract states that R2BC outperforms centralized BC by 3.25x (navigation) and 5.9x (block pushing). In the navigation task, the zero-intervention difference is reported as not statistically significant (p = 0.0773, Section VII-B), with significance only emerging after allowing 1–2 interventions. The 3.25x figure is therefore an overstatement when presented without qualification in the abstract. The authors do acknowledge the lack of significance in the body, which is good, but the abstract should either include the caveat, report the p-value, or drop the navigation multiplier and present the significant intervention results instead. This matters because the abstract is the primary summary of the paper's contributions.
minor comments (6)
  1. [Figure 3] The y-axis is labeled 'Loss Gap (Log)'—please specify the base of the log and define whether the gap is absolute or relative. Also clarify the training set used for computing training loss in the online methods (e.g., which demonstration buffer).
  2. [Algorithm 1, line 15] The update condition 'if c mod k = 0' means that when k=1, updates occur after every round, but when k>1, the first update occurs after the first round because c=0. This is fine, but it would be clearer to state that c counts full round-robin passes, and that updates occur at the end of every k-th pass. Also clarify whether all agents are updated simultaneously using all data collected so far or only the latest k episodes.
  3. [Section VII-A] The sentence 'the authors provided two sets of 240 demonstrations for each task' is ambiguous. Does this mean 240 demonstrations per method per task (480 total per task), or 240 total split between methods? Please clarify. Also note that the human demonstrators are authors; this is acceptable for a proof of concept, but a brief discussion of potential bias and a plan for a user study with naive subjects would strengthen the paper.
  4. [Section VI, Figure 4] The 'Offline R2BC (No-Op Agents)' variant assumes that non-demonstrating agents execute no action. In many multi-agent environments, a 'no-op' may not be a valid action or may have a different effect than standing still. Please specify how no-op actions were implemented and whether the environment supports them.
  5. [Abstract and Introduction] The claim 'we are the first to propose and deploy a behavior cloning method for multi-agent systems that learns solely from online single-agent demonstrations' is strong. While the related work appears thorough, consider softening to 'to the best of our knowledge' in the abstract as well, and ensure that the related work section explicitly contrasts R2BC with Fleet-DAgger and other interactive IL methods, which the paper already does but could be expanded.
  6. [General] There are minor formatting issues, such as 'Round-RobinBehaviorCloning' in the introduction and inconsistent spacing in the figure captions. A careful proofread is recommended.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: R2BC's performance is evaluated against external oracle baselines and held-out environment reward, not against its own fitted quantities.

full rationale

I examined the paper's derivation chain: the problem formulation defines single-agent demonstrations (Eq. 2), Algorithm 1 constructs R2BC using a standard behavior-cloning loss, and the experiments compare R2BC to JBC, DART, and DAgger using labels from a shared MAPPO oracle. No step defines a predicted quantity in terms of a fitted parameter, and no claimed result reduces by construction to the method's inputs. The oracle MAPPO policy is an external label source used by all baselines, so R2BC's success is not the same quantity it was trained on. The real-robot comparison uses human-provided JBC demonstrations that are difficult to produce, but that is an experimental limitation or confound, not a definitional identity. Self-citations in the paper ([12], [23]) appear only as background for single-agent IL methods and robot hardware; they are not load-bearing for the R2BC claim. The conclusion's admission that convergence and covariate-shift guarantees are absent ('developing theoretical guarantees for convergence and covariate shift reduction would further reinforce the intuition provided in this work') is an acknowledged limitation, not evidence of circularity. No circular step meeting the required quote-and-reduction standard was found.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

R2BC introduces no new physical or latent entities; it is a data-collection and training protocol over existing policies and buffers. The main load-bearing assumptions are the infeasibility of joint human demonstrations, the validity of the MAPPO oracle labeler in simulation, and convergence of the online co-adaptation loop, which is admitted as unproved.

free parameters (3)
  • Update frequency k
    Policies are updated every k demonstrations; k is not specified or swept in the main text, though it is a control knob of the algorithm.
  • Per-task demonstration budget
    Figure 2 uses different demonstration counts per task; no principled selection rule is given, which affects how R2BC compares with joint-action baselines.
  • BC hyperparameters (learning rate, minibatch) = lr=1e-3, batch=256
    Reported in Section V-A but not justified or ablated; standard for behavior cloning.
assumptions (4)
  • domain assumption A single human cannot provide synchronized near-optimal joint demonstrations for N>=2 robots.
    Motivates the problem formulation; empirically reasonable for many real tasks but not established as a hard bound.
  • domain assumption The MAPPO policy trained on ground-truth reward is a valid oracle demonstrator for both joint and single-agent action labels.
    The synthetic evaluation depends on this oracle; real-human experiments partially support the setting but do not validate the oracle assumption.
  • domain assumption MSE behavior cloning on per-agent buffers is a sufficient learning rule; compounding error is handled by the online round-robin data distribution.
    R2BC's mechanism relies on this; the paper provides empirical evidence but no distribution-shift or convergence proof.
  • ad hoc to paper The co-adaptation of teammate policies during round-robin training converges to a stable and useful team policy.
    Not proven; the authors explicitly list convergence guarantees as future work in the conclusion, making this the load-bearing unproved premise.

how reviews work

0 comments
Cite this review

Pith. "Pith review of R2BC: Multi-Agent Imitation Learning from Single-Agent Demonstrations." pith.science (2026). https://pith.science/paper/6SNHEQ6Z

@misc{pith2026251018085,
  author       = {Pith},
  title        = {Pith review of: R2BC: Multi-Agent Imitation Learning from Single-Agent Demonstrations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6SNHEQ6Z}},
  note         = {Machine review of arXiv:2510.18085}
}
read the original abstract

Imitation Learning (IL) is a natural way for humans to teach robots, particularly when high-quality demonstrations are easy to obtain. While IL has been widely applied to single-robot settings, relatively few studies have addressed the extension of these methods to multi-agent systems, especially in settings where a single human must provide demonstrations to a team of collaborating robots. In this paper, we introduce and study Round-Robin Behavior Cloning (R2BC), a method that enables a single human operator to effectively train multi-robot systems through sequential, single-agent demonstrations. Our approach allows the human to teleoperate one agent at a time and incrementally teach multi-agent behavior to the entire system, without requiring demonstrations in the joint multi-agent action space. We show that R2BC methods match, and in some cases surpass, the performance of an oracle behavior cloning approach trained on privileged synchronized demonstrations across four multi-agent simulated tasks. Finally, we deploy R2BC on two physical robot tasks trained using real human demonstrations.

Figures

Figures reproduced from arXiv: 2510.18085 by the authors.

Figure 1
Figure 1. Round-Robin Behavior Cloning (R2BC): Traditional Behavior Cloning (left) requires coordinated and centralized demonstrations, where an expert demonstrates actions near-optimally for all agents. In multi-agent domains, a lone human operator may not be able to provide high-quality demonstrations due to underactuated control and increased cognitive burden. Our method (right), R2BC, removes this restriction by letting t… view at source ↗
Figure 2
Figure 2. Simulation Results: We compare the R2BC method to a set of three baselines that assume oracle (privileged) access to a joint action demonstrator. Results are averaged over 10 seeds with shaded regions indicating a 99% confidence interval. Top: Four multi-agent tasks selected from the Vectorized Multi-Agent Simulator (VMAS) [3] including Navigation (N = 3), Balance (N = 3), Buzz Wire (N = 2), and Transport (N = 3). B… view at source ↗
Figure 4
Figure 4. R2BC Ablations: Comparison of R2BC to several ablations across four simulated environments. Across all environments, R2BC consistently outperforms the two offline variants and matches or exceeds the oracle JBC methods. Error bars indicate 99% confidence interval over 10 seeded runs. to near-optimal, coordinated demonstrations—surprisingly, we found that in all simulated domains, R2BC outperforms or achieves parity w… view at source ↗
Figures from the paper (2 more)
Figure 5
Figure 5. Figure 5: Real Robot Experiments: (a) The HeRo+ Robot, a differential drive mobile robot that acts using a forward velocity and angular velocity, (b) AruCo markers used to track robot/goal states, (c) R2BC demonstrations are provided in simulation by a real human, (d) a real wor…
Figure 6
Figure 6. Figure 6: Real Robot Results: JBC and R2BC policies trained from real human demonstrations in simulation are deployed directly onto the real robots in two tasks. To account for sim-to-real distribution shift, we report results with 0, 1, and 2 online interventions, where a human…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

  1. Duet: Dual-Robot Understanding via Efficient Teaching

    cs.RO 2026-06 unverdicted novelty 5.0 of 10

    DUET pretrains collaborative policies on human-human VR demonstrations then fine-tunes on minimal robot teleoperation data, achieving equal or better performance than robot-only baselines with 5.4x faster collection a...

Reference graph

Works this paper leans on

38 extracted references · 3 linked inside Pith · cited by 1 Pith paper

  1. [1]

    Decentralized control of a heterogeneous human–robot team for exploration and patrolling.IEEE Transactions on Automation Science and Engineering, 2021

    Marco Aggravi, Giuseppe Sirignano, Paolo Robuffo Giordano, and Claudio Pacchierotti. Decentralized control of a heterogeneous human–robot team for exploration and patrolling.IEEE Transactions on Automation Science and Engineering, 2021

    2021

  2. [2]

    Mapf-gpt: Imitation learning for multi-agent pathfinding at scale

    Anton Andreychuk, Konstantin Yakovlev, Aleksandr Panov, and Alexey Skrynnik. Mapf-gpt: Imitation learning for multi-agent pathfinding at scale. InProceedings of the AAAI Conference on Artificial Intelligence, 2025

    2025

  3. [3]

    Vmas: A vectorized multi-agent simulator for collective robot learning

    Matteo Bettini, Ryan Kortvelesy, Jan Blumenkamp, and Amanda Prorok. Vmas: A vectorized multi-agent simulator for collective robot learning. InInternational Symposium on Distributed Autonomous Robotic Systems. Springer, 2022

    2022

  4. [4]

    Multi-agent imitation learning for driving simulation

    Raunak P Bhattacharyya, Derek J Phillips, Blake Wulfe, Jeremy Mor- ton, Alex Kuefler, and Mykel J Kochenderfer. Multi-agent imitation learning for driving simulation. In2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2018

    2018

  5. [5]

    Adaptive task planning for multi-robot smart warehouse.Ieee Access, 2021

    Ali Bolu and ¨Omer Korc ¸ak. Adaptive task planning for multi-robot smart warehouse.Ieee Access, 2021

    2021

  6. [6]

    Transformer-based imitative rein- forcement learning for multirobot path planning.IEEE Transactions on Industrial Informatics, 2023

    Lin Chen, Yaonan Wang, Zhiqiang Miao, Yang Mo, Mingtao Feng, Zhen Zhou, and Hesheng Wang. Transformer-based imitative rein- forcement learning for multirobot path planning.IEEE Transactions on Industrial Informatics, 2023

    2023

  7. [7]

    Survey of imitation learning for robotic manipulation

    Bin Fang, Shidong Jia, Di Guo, Muhua Xu, Shuhuan Wen, and Fuchun Sun. Survey of imitation learning for robotic manipulation. International Journal of Intelligent Robotics and Applications, 2019

    2019

  8. [8]

    Multi-agent generative adversarial interactive self- imitation learning for auv formation control and obstacle avoidance

    Zheng Fang, Tianhao Chen, Tian Shen, Dong Jiang, Zheng Zhang, and Guangliang Li. Multi-agent generative adversarial interactive self- imitation learning for auv formation control and obstacle avoidance. IEEE Robotics and Automation Letters, 2025

    2025

Show all 38 references
  1. [9]

    Teleoperation of multi- robot and multi-property systems

    Ildar Farkhatdinov and Jee-Hwan Ryu. Teleoperation of multi- robot and multi-property systems. In2008 6th IEEE International Conference on Industrial Informatics. IEEE, 2008

    2008

  2. [10]

    Multi-agent imitation learning based energy management of a microgrid with hybrid energy storage and real-time pricing.IEEE Internet of Things Journal, 2025

    Shuhua Gao, Yizhuo Xu, Zhaoqian Zhang, Zhengfang Wang, Xiaoyu Zhou, and Jing Wang. Multi-agent imitation learning based energy management of a microgrid with hybrid energy storage and real-time pricing.IEEE Internet of Things Journal, 2025

    2025

  3. [11]

    Applica- tion of multi-robot systems to disaster-relief scenarios with limited communication

    Jason Gregory, Jonathan Fink, Ethan Stump, Jeffrey Twigg, John Rogers, David Baran, Nicholas Fung, and Stuart Young. Applica- tion of multi-robot systems to disaster-relief scenarios with limited communication. InField and service robotics: results of the 10th international co...

    2016

  4. [12]

    Thriftydagger: Budget-aware novelty and risk gating for interactive imitation learning

    Ryan Hoque, Ashwin Balakrishna, Ellen Novoseller, Albert Wilcox, Daniel S Brown, and Ken Goldberg. Thriftydagger: Budget-aware novelty and risk gating for interactive imitation learning. InConference on Robot Learning, 2021

    2021

  5. [13]

    Fleet-dagger: Interactive robot fleet learning with scalable human supervision

    Ryan Hoque, Lawrence Yunliang Chen, Satvik Sharma, Karthik Dharmarajan, Brijen Thananjeyan, Pieter Abbeel, and Ken Goldberg. Fleet-dagger: Interactive robot fleet learning with scalable human supervision. InConference on Robot Learning, pages 368–380. PMLR, 2023

    2023

  6. [14]

    Multi-agent decision-making at unsignalized inter- sections with reinforcement learning from demonstrations

    Chang Huang, Junqiao Zhao, Hongtu Zhou, Hai Zhang, Xiao Zhang, and Chen Ye. Multi-agent decision-making at unsignalized inter- sections with reinforcement learning from demonstrations. InIEEE Intelligent Vehicles Symposium (IV), 2023

    2023

  7. [15]

    From one to many: How active robot swarm sizes influence human cognitive processes

    Julian Kaduk, M ¨uge Cavdan, Knut Drewing, and Heiko Hamann. From one to many: How active robot swarm sizes influence human cognitive processes. In2024 33rd IEEE International Conference on Robot and Human Interactive Communication (ROMAN). IEEE, 2024

    2024

  8. [16]

    Dart: Noise injection for robust imitation learning

    Michael Laskey, Jonathan Lee, Roy Fox, Anca Dragan, and Ken Goldberg. Dart: Noise injection for robust imitation learning. In Conference on robot learning. PMLR, 2017

    2017

  9. [17]

    Coordi- nated multi-agent imitation learning

    Hoang M Le, Yisong Yue, Peter Carr, and Patrick Lucey. Coordi- nated multi-agent imitation learning. InInternational Conference on Machine Learning. PMLR, 2017

    2017

  10. [18]

    A two-stage imitation learning framework for the multi-target search problem in swarm robotics.Neurocomputing, 2019

    Jie Li and Ying Tan. A two-stage imitation learning framework for the multi-target search problem in swarm robotics.Neurocomputing, 2019

    2019

  11. [19]

    Collective behavior clone with visual attention via neural interaction graph prediction.arXiv preprint arXiv:2503.06869, 2025

    Kai Li, Zhao Ma, Liang Li, and Shiyu Zhao. Collective behavior clone with visual attention via neural interaction graph prediction.arXiv preprint arXiv:2503.06869, 2025

    2025 arXiv

  12. [20]

    Controlling assistive robots with learned latent actions

    Dylan P Losey, Krishnan Srinivasan, Ajay Mandlekar, Animesh Garg, and Dorsa Sadigh. Controlling assistive robots with learned latent actions. In2020 IEEE International Conference on Robotics and Automation (ICRA), 2020

    2020

  13. [21]

    Learn to behave! rapid training of behavior automata

    Sean Luke and Vittorio Amos Ziparo. Learn to behave! rapid training of behavior automata. InProceedings of adaptive and learning agents workshop at aamas. Citeseer, 2010

    2010

  14. [22]

    Inverse factorized soft q-learning for cooperative multi-agent imitation learning.Advances in Neural Information Processing Systems, 2024

    Tien Mai, Thanh Nguyen, et al. Inverse factorized soft q-learning for cooperative multi-agent imitation learning.Advances in Neural Information Processing Systems, 2024

    2024

  15. [23]

    Discovery and deployment of emergent robot swarm behaviors via representation learning and real2sim2real transfer

    Connor Mattson, Varun Raveendra, Ricardo Vega, Cameron Nowzari, Daniel S Drew, and Daniel S Brown. Discovery and deployment of emergent robot swarm behaviors via representation learning and real2sim2real transfer. InProceedings of the 24th International Conference on Autonomou...

    2025

  16. [24]

    Transparency in multi-human multi-robot interaction.arXiv preprint arXiv:2101.10495, 2021

    Jayam Patel, Tyagaraja Ramaswamy, Zhi Li, and Carlo Pinciroli. Transparency in multi-human multi-robot interaction.arXiv preprint arXiv:2101.10495, 2021

    2021 arXiv

  17. [25]

    Diffusion- based multi-agent reinforcement learning with communication

    Xinyue Qi, Jianhang Tang, Jiangming Jin, and Yang Zhang. Diffusion- based multi-agent reinforcement learning with communication. InAsia Pacific Wireless Communications Symposium, 2024

    2024

  18. [26]

    Collaborative multi- robot search and rescue: Planning, coordination, perception, and active vision.Ieee Access, 2020

    Jorge Pena Queralta, Jussi Taipalmaa, Bilge Can Pullinen, Vic- tor Kathan Sarker, Tuan Nguyen Gia, Hannu Tenhunen, Moncef Gabbouj, Jenni Raitoharju, and Tomi Westerlund. Collaborative multi- robot search and rescue: Planning, coordination, perception, and active vision.Ieee Ac...

    2020

  19. [27]

    Hero 2.0: A low-cost robot for swarm robotics research

    Paulo Rezeck, H ´ector Azp ´urua, Mauricio FS Correa, and Luiz Chaimowicz. Hero 2.0: A low-cost robot for swarm robotics research. Autonomous Robots, 2023

    2023

  20. [28]

    Efficient reductions for imitation learning

    St ´ephane Ross and Drew Bagnell. Efficient reductions for imitation learning. InInternational conference on artificial intelligence and statistics, 2010

    2010

  21. [29]

    A reduction of imitation learning and structured prediction to no-regret online learning

    St ´ephane Ross, Geoffrey Gordon, and Drew Bagnell. A reduction of imitation learning and structured prediction to no-regret online learning. InInternational Conference on Artificial Intelligence and Statistics, 2011

    2011

  22. [30]

    Multi-robot persistent surveil- lance with connectivity constraints.IEEE Access, 2020

    J ¨urgen Scherer and Bernhard Rinner. Multi-robot persistent surveil- lance with connectivity constraints.IEEE Access, 2020

    2020

  23. [31]

    Proximal policy optimization algorithms.arXiv preprint arXiv:1707.06347, 2017

    John Schulman, Filip Wolski, Prafulla Dhariwal, Alec Radford, and Oleg Klimov. Proximal policy optimization algorithms.arXiv preprint arXiv:1707.06347, 2017

    2017 arXiv

  24. [32]

    Multi- agent generative adversarial imitation learning.Advances in neural information processing systems, 2018

    Jiaming Song, Hongyu Ren, Dorsa Sadigh, and Stefano Ermon. Multi- agent generative adversarial imitation learning.Advances in neural information processing systems, 2018

    2018

  25. [33]

    Lfd training of heterogeneous formation behaviors

    William G Squires and Sean Luke. Lfd training of heterogeneous formation behaviors. InAAAI Spring Symposia, 2018

    2018

  26. [34]

    Fast multi-robot motion planning via imitation learning of mixed-integer programs.IFAC-PapersOnLine, 2021

    Mohit Srinivasan, Ankush Chakrabarty, Rien Quirynen, Nobuyuki Yoshikawa, Toshisada Mariyama, and Stefano Di Cairano. Fast multi-robot motion planning via imitation learning of mixed-integer programs.IFAC-PapersOnLine, 2021

    2021

  27. [35]

    Multi-agent imitation learning: Value is easy, regret is hard.Advances in Neural Information Processing Systems, 2024

    Jingwu Tang, Gokul Swamy, Fei Fang, and Steven Z Wu. Multi-agent imitation learning: Value is easy, regret is hard.Advances in Neural Information Processing Systems, 2024

    2024

  28. [36]

    Reducing cog- nitive load in teleoperating swarms of robots through a data-driven shared control approach

    Enrico Turco, Chiara Castellani, Valerio Bo, Claudio Pacchierotti, Domenico Prattichizzo, and Tommaso Lisini Baldi. Reducing cog- nitive load in teleoperating swarms of robots through a data-driven shared control approach. In2024 IEEE/RSJ International Conference on Intelligen...

    2024

  29. [37]

    The surprising effectiveness of ppo in cooperative multi-agent games.Advances in neural information processing systems, 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, 2022

    2022

  30. [38]

    Clone swarms: Learning to predict and control multi-robot systems by imitation

    Siyu Zhou, Mariano J Phielipp, Jorge A Sefair, Sara I Walker, and Heni Ben Amor. Clone swarms: Learning to predict and control multi-robot systems by imitation. In2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2019

    2019

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.