mlsanityData to 14 Oct 2026

PIMbot: A Self-Adaptive Attack Framework for Adversarial Manipulation of Multi-Robot Reinforcement Learning

arXiv 2605.23027 ↗First posted 21 May 2026cs.ROcited by 1

On the map: Deep reinforcement learning methods → Adversarial robustness and RL in finance

Abstract (arXiv)

Recent research has demonstrated the potential of reinforcement learning in effective multi-robot collaboration, particularly in social dilemmas where robots face a trade-off between self-interest and collective benefits. However, environmental factors such as miscommunication and adversarial robots can impact cooperation, making it crucial to explore how multi-robot communication can be manipulated to achieve different outcomes. This paper presents PIMbot, a framework that manipulates outcomes via two complementary levers: (i) incentive manipulation of the reward channel and (ii) policy manipulation of an agent's own actions. An adaptive multi-objective controller balances these levers in an online manner. Our work introduces a novel approach to manipulation in recent multi-agent RL social dilemmas that utilize a unique reward function for incentivization. By utilizing our proposed PIMbot mechanisms, a robot is able to manipulate the social dilemma environment effectively. Comprehensive experimental results demonstrate the effectiveness of our proposed methods in the Gazebo-simulated multi-robot environment. Moreover, a real embedded device case study on NVIDIA Jetson Orin Nano quantifies system cost and validates PIMbot's effectiveness on realistic autonomous embedded systems scenarios beyond simulation. Together, these results position PIMbot as a rigorous stress-test tool exposing critical vulnerabilities in multi-robot cooperative tasks.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in Adversarial robustness and RL in finance, one of 15 directions in Deep reinforcement learning methods. Builds on 48 of its 75 references that are in the map. First posted on arXiv 21 May 2026.

Builds on

48 of 75 references.

References · 48

  1. 2018QMIX: Monotonic Value Function Factorisation for Deep Multi-Agent Reinforcement Learningcited by 708
  2. 2019The StarCraft Multi-Agent Challengecited by 597
  3. 2016Learning to Communicate with Deep Multi-Agent Reinforcement Learningcited by 575
  4. 2019Multi-Agent Reinforcement Learning: A Selective Overview of Theories and Algorithmscited by 567
  5. 2017Value-Decomposition Networks For Cooperative Multi-Agent Learningcited by 503
  6. 2019QTRAN: Learning to Factorize with Transformation for Cooperative Multi-Agent Reinforcement Learningcited by 379
  7. 2016Learning Multiagent Communication with Backpropagationcited by 366
  8. 2020QPLEX: Duplex Dueling Multi-Agent Q-Learningcited by 270
  9. 2017Learning with Opponent-Learning Awarenesscited by 222
  10. 2020PettingZoo: Gym for Multi-Agent Reinforcement Learningcited by 220
  11. 2019Adversarial Policies: Attacking Deep Reinforcement Learningcited by 214
  12. 2018Mean Field Multi-Agent Reinforcement Learningcited by 207
36 more
  1. 2019MAVEN: Multi-Agent Variational Explorationcited by 185
  2. 2018Learning when to Communicate at Scale in Multiagent Cooperative and Competitive Taskscited by 121
  3. 2018On Learning Intrinsic Rewards for Policy Gradient Methodscited by 102
  4. 2019Dealing with Non-Stationarity in Multi-Agent Deep Reinforcement Learningcited by 93
  5. 2020Learning to Incentivize Other Learning Agentscited by 57
  6. 2022NICGSlowDown: Evaluating the Efficiency Robustness of Neural Image Caption Generation Modelscited by 50
  7. 2021Adversarial Attacks On Multi-Agent Communicationcited by 49
  8. 2025Dynamic Benchmarking of Reasoning Capabilities in Code Large Language Models Under Data Contaminationcited by 23
  9. 2022What is the Solution for State-Adversarial Multi-Agent Reinforcement Learning?cited by 22
  10. 2021Adaptive Incentive Design with Multi-Agent Meta-Gradient Reinforcement Learningcited by 19
  11. 2023Attacking Cooperative Multi-Agent Reinforcement Learning by Adversarial Minority Influencecited by 17
  12. 2023White-Box Multi-Objective Adversarial Attack on Dialogue Generationcited by 14
  13. 2023Dynamic Transformers Provide a False Sense of Efficiencycited by 14
  14. 2024TTSlow: Slow Down Text-to-Speech with Efficiency Robustness Evaluationscited by 13
  15. 2023RT-LM: Uncertainty-Aware Resource Management for Real-Time Inference of Language Modelscited by 11
  16. 2023Safety Alignment in NLP Tasks: Weakly Aligned Summarization as an In-Context Attackcited by 11
  17. 2023Efficient Adversarial Attacks on Online Multi-agent Reinforcement Learningcited by 10
  18. 2021Autonomous Teamed Exploration of Subterranean Environments using Legged and Aerial Robotscited by 7
  19. 2022DC-MRTA: Decentralized Multi-Robot Task Allocation and Navigation in Complex Environmentscited by 7
  20. 2025Your Compiler is Backdooring Your Model: Understanding and Exploiting Compilation Inconsistency Vulnerabilities in Deep Learning Compilerscited by 7
  21. 2021Meeting-Merging-Mission: A Multi-robot Coordinate Framework for Large-Scale Communication-Limited Explorationcited by 6
  22. 2023PIMbot: Policy and Incentive Manipulation for Multi-Robot Reinforcement Learning in Social Dilemmascited by 6
  23. 2024Unveiling the Achilles' Heel of NLG Evaluators: A Unified Adversarial Framework Driven by Large Language Modelscited by 6
  24. 2022Dynamic Collaborative Multi-Agent Reinforcement Learning Communication for Autonomous Drone Reforestationcited by 5
  25. 2023Robust Communicative Multi-Agent Reinforcement Learning with Active Defensecited by 5
  26. 2021Multi-Robot Task Planning under Individual and Collaborative Temporal Logic Specificationscited by 4
  27. 2022Traversing Supervisor Problem: An Approximately Optimal Approach to Multi-Robot Assistancecited by 4
  28. 2024Transferable Adversarial Attacks against ASRcited by 4
  29. 2024Reciprocal Reward Influence Encourages Cooperation From Self-Interested Agentscited by 3
  30. 2025Dr Genre: Reinforcement Learning from Decoupled LLM Feedback for Generic Text Rewritingcited by 3
  31. 2025LeMix: Unified Scheduling for LLM Training and Inference on Multi-GPU Systemscited by 3
  32. 2019An Efficient Scheduling Algorithm for Multi-Robot Task Allocation in Assembling Aircraft Structurescited by 2
  33. 2025MACE: A Hybrid LLM Serving System with Colocated SLO-aware Continuous Retraining Alignmentcited by 2
  34. 2025Enabling Low-Cost Secure Computing on Untrusted In-Memory Architecturescited by 1
  35. 2025A Near-Cache Architectural Framework for Cryptographic Computingcited by 1
  36. 2025SAIL: SRAM-Accelerated LLM Inference System with Lookup-Table-based GEMVcited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Boosting Transferable Adversarial Attacks against Deep Reinforcement Learningcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025RoBCtrl: Attacking GNN-Based Social Bot Detectors via Reinforced Manipulation of Bots Control Interaction
  2. 2024SUB-PLAY: Adversarial Policies against Partially Observed Multi-Agent Reinforcement Learning Systems
  3. 2024RAT: Adversarial Attacks on Deep Reinforcement Agents for Targeted Behaviors
  4. 2025Constrained Black-Box Attacks Against Cooperative Multi-Agent Reinforcement Learning
  5. 2025HARL-A: An Extensible Benchmark Framework for Heterogeneous Multi-Agent Adversarial Reinforcement Learning in IsaacLab
  6. 2026Interaction-Breaking Adversarial Learning Framework for Robust Multi-Agent Reinforcement Learning
  7. 2026ALSO: Adversarial Online Strategy Optimization for Social Agents
  8. 2024FightLadder: A Benchmark for Competitive Multi-Agent Reinforcement Learning