ACPO: Agent-Chained Policy Optimization for Multi-Agent Reinforcement Learning

• URL: http://arxiv.org/abs/2606.30072v1
• Date: Mon, 29 Jun 2026 10:04:15 GMT
• Title: ACPO: Agent-Chained Policy Optimization for Multi-Agent Reinforcement Learning
• Abstract summary: Multi-Agent Reinforcement Learning requires agents to collectively maximize a shared return.<n>Prior methods largely follow two approaches: independent factorized updates with centralized critics, or alternating best-response updates.<n>We show the joint policy gradient admits an exact decentralized decomposition of per-agent terms, each formed from per-agent score functions and decentralized critics.<n>Central to this result is a serialized view of the simultaneous joint decision in which agents commit actions one at a time, each conditioning on a belief over preceding actions.
• Score: 47.96591344642519
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Cooperative tasks in Multi-Agent Reinforcement Learning (MARL) require agents to collectively maximize a shared return. Under the Centralized Training with Decentralized Execution (CTDE) paradigm, policy gradients have remained difficult to compute directly. Prior methods largely follow two approaches: independent factorized updates with centralized critics, which lack general joint-improvement guarantees without value decomposition assumptions, or alternating best-response updates, which can converge to suboptimal Nash Equilibria. In this paper, we show the joint policy gradient admits an exact decentralized decomposition of per-agent terms, each formed from per-agent score functions and decentralized critics. Based on this decomposition, we develop Agent-Chained Policy Optimization (ACPO), where actors are trained independently, with their updates together constituting a single step on the joint policy gradient. Central to this result is a serialized view of the simultaneous joint decision in which agents commit actions one at a time, each conditioning on a belief over preceding actions. The belief acts as the coordination mechanism which ties the independent per-agent updates into a joint gradient step. We evaluate ACPO on Multi-Robot Warehouse, SMACv2, and MA-MuJoCo, where it outperforms strong baselines, with the gap widening as the number of agents grows.

Related papers

• GRASP: Gradient Realignment via Active Shared Perception for Multi-Agent Collaborative Optimization [41.52202306408042]
  Non-stationarity arises from concurrent policy updates and leads to persistent environmental fluctuations.<n>We propose Realignment via Active Shared Perception (GRASP), a novel framework that defines generalized Bellman equilibrium as a stable objective for policy evolution.
  arXiv  Detail & Related papers  (2026-04-01T10:26:22Z)
• Decentralized Policy Optimization [21.59254848913971]
  We propose textitdecentralized policy optimization (DPO), a decentralized actor-critic algorithm with monotonic improvement and convergence guarantee. Empirically, we compare DPO with IPPO in a variety of cooperative multi-agent tasks, covering discrete and continuous action spaces, and fully and partially observable environments.
  arXiv  Detail & Related papers  (2022-11-06T05:38:23Z)
• Monotonic Improvement Guarantees under Non-stationarity for Decentralized PPO [66.5384483339413]
  We present a new monotonic improvement guarantee for optimizing decentralized policies in cooperative Multi-Agent Reinforcement Learning (MARL) We show that a trust region constraint can be effectively enforced in a principled way by bounding independent ratios based on the number of agents in training.
  arXiv  Detail & Related papers  (2022-01-31T20:39:48Z)
• Iterated Reasoning with Mutual Information in Cooperative and Byzantine Decentralized Teaming [0.0]
  We show that reformulating an agent's policy to be conditional on the policies of its teammates inherently maximizes Mutual Information (MI) lower-bound when optimizing under Policy Gradient (PG) Our approach, InfoPG, outperforms baselines in learning emergent collaborative behaviors and sets the state-of-the-art in decentralized cooperative MARL tasks.
  arXiv  Detail & Related papers  (2022-01-20T22:54:32Z)
• Dealing with Non-Stationarity in Multi-Agent Reinforcement Learning via Trust Region Decomposition [52.06086375833474]
  Non-stationarity is one thorny issue in multi-agent reinforcement learning. We introduce a $delta$-stationarity measurement to explicitly model the stationarity of a policy sequence. We propose a trust region decomposition network based on message passing to estimate the joint policy divergence.
  arXiv  Detail & Related papers  (2021-02-21T14:46:50Z)
• Multi-Agent Trust Region Policy Optimization [34.91180300856614]
  We show that the policy update of TRPO can be transformed into a distributed consensus optimization problem for multi-agent cases. We propose a decentralized MARL algorithm, which we call multi-agent TRPO (MATRPO)
  arXiv  Detail & Related papers  (2020-10-15T17:49:47Z)
• F2A2: Flexible Fully-decentralized Approximate Actor-critic for Cooperative Multi-agent Reinforcement Learning [110.35516334788687]
  Decentralized multi-agent reinforcement learning algorithms are sometimes unpractical in complicated applications. We propose a flexible fully decentralized actor-critic MARL framework, which can handle large-scale general cooperative multi-agent setting. Our framework can achieve scalability and stability for large-scale environment and reduce information transmission.
  arXiv  Detail & Related papers  (2020-04-17T14:56:29Z)
• FACMAC: Factored Multi-Agent Centralised Policy Gradients [103.30380537282517]
  We propose FACtored Multi-Agent Centralised policy gradients (FACMAC) It is a new method for cooperative multi-agent reinforcement learning in both discrete and continuous action spaces. We evaluate FACMAC on variants of the multi-agent particle environments, a novel multi-agent MuJoCo benchmark, and a challenging set of StarCraft II micromanagement tasks.
  arXiv  Detail & Related papers  (2020-03-14T21:29:09Z)
• Counterfactual Multi-Agent Policy Gradients [47.45255170608965]
  We propose a new multi-agent actor-critic method called counterfactual multi-agent (COMA) policy gradients.<n>COMA uses a centralised critic to estimate the Q-function and decentralised actors to optimise the agents' policies.<n>We evaluate COMA in the testbed of StarCraft unit micromanagement, using a decentralised variant with significant partial observability.
  arXiv  Detail & Related papers  (2017-05-24T18:52:17Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.