Learning Partial Action Replacement in Offline MARL

• URL: http://arxiv.org/abs/2603.28573v1
• Date: Mon, 30 Mar 2026 15:28:13 GMT
• Title: Learning Partial Action Replacement in Offline MARL
• Abstract summary: Partial Action Replacement mitigates this by anchoring a subset of agents to dataset actions.<n>We introduce PLCQL, a framework that formulates PAR subset selection as a contextual bandit problem.<n>We prove a value-error bound showing that the estimation error scales linearly with the expected number of deviating agents.
• Score: 11.861550409939818
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Offline multi-agent reinforcement learning (MARL) faces a critical challenge: the joint action space grows exponentially with the number of agents, making dataset coverage exponentially sparse and out-of-distribution (OOD) joint actions unavoidable. Partial Action Replacement (PAR) mitigates this by anchoring a subset of agents to dataset actions, but existing approach relies on enumerating multiple subset configurations at high computational cost and cannot adapt to varying states. We introduce PLCQL, a framework that formulates PAR subset selection as a contextual bandit problem and learns a state-dependent PAR policy using Proximal Policy Optimisation with an uncertainty-weighted reward. This adaptive policy dynamically determines how many agents to replace at each update step, balancing policy improvement against conservative value estimation. We prove a value-error bound showing that the estimation error scales linearly with the expected number of deviating agents. Compared with the previous PAR-based method SPaCQL, PLCQL reduces the number of per-iteration Q-function evaluations from n to 1, significantly improving computational efficiency. Empirically, PLCQL achieves the highest normalised scores on 66% of tasks across MPE, MaMuJoCo, and SMAC benchmarks, outperforming SPaCQL on 84% of tasks while substantially reducing computational cost.

Related papers

• Adaptive TD-Lambda for Cooperative Multi-agent Reinforcement Learning [28.30054999007724]
  Some recent works link the adaptive $$ value to the policy distribution in the single-agent reinforcement learning area.<n>Because of the large joint action space from multiple number of agents, and the limited transition data in Multi-agent Reinforcement Learning, the policy distribution is infeasible to be calculated statistically.<n>We employ a parametric likelihood-free density ratio estimator with two replay buffers instead of calculating statistically.
  arXiv  Detail & Related papers  (2026-05-12T09:56:24Z)
• Agent psychometrics: Task-level performance prediction in agentic coding benchmarks [24.348135523715815]
  We present a framework for predicting success or failure on individual tasks tailored to the agentic coding regime.<n>Our approach augments Item Response Theory (IRT) with rich features extracted from tasks, including issue statements, repository contexts, solutions, and test cases.
  arXiv  Detail & Related papers  (2026-04-01T07:59:59Z)
• Improving Search Agent with One Line of Code [68.58667107354253]
  Tool-based Agentic Reinforcement Learning (TARL) has emerged as a promising paradigm for training search agents to interact with external tools.<n>We propose textbfSearch textbfAgent textbfPolicy textbfOptimization (textbfSAPO), which stabilizes training via a conditional token-level KL constraint.
  arXiv  Detail & Related papers  (2026-03-10T04:07:39Z)
• Unbiased Dynamic Pruning for Efficient Group-Based Policy Optimization [60.87651283510059]
  Group Relative Policy Optimization (GRPO) effectively scales LLM reasoning but incurs prohibitive computational costs.<n>We propose Dynamic Pruning Policy Optimization (DPPO), a framework that enables dynamic pruning while preserving unbiased gradient estimation.<n>To mitigate the data sparsity induced by pruning, we introduce Dense Prompt Packing, a window-based greedy strategy.
  arXiv  Detail & Related papers  (2026-03-04T14:48:53Z)
• Partial Action Replacement: Tackling Distribution Shift in Offline MARL [11.861550409939818]
  offline multi-agent reinforcement learning (MARL) is severely hampered by the challenge of evaluating out-of-distribution joint actions.<n>We develop Soft-Partial Conservative Q-Learning (SPaCQL) to mitigate OOD issue and dynamically weighting different PAR strategies.<n>Our theoretical results also indicate that SPaCQL adaptively addresses distribution shift using uncertainty-informed weights.
  arXiv  Detail & Related papers  (2025-11-10T20:56:58Z)
• Improving monotonic optimization in heterogeneous multi-agent reinforcement learning with optimal marginal deterministic policy gradient [18.64288030584699]
  heterogeneous multi-agent reinforcement learning (MARL)<n>Objectively replace the sequentially computed $Q_psi*(s,a_1:i)$ with the Optimal Marginal Q function $phi_psi*(s,a_1:i)$ derived from Q-functions.<n>Generalized Q Critic (GQC) as the critic function, employing pessimistic uncertainty-constrained loss to optimize different Q-value estimations.
  arXiv  Detail & Related papers  (2025-07-14T07:16:01Z)
• Simulation-Based Optimistic Policy Iteration For Multi-Agent MDPs with Kullback-Leibler Control Cost [3.9052860539161918]
  This paper proposes an agent-based optimistic policy (OPI) scheme for learning stationary optimal policies in Markov Decision Processes (MDPs) The proposed scheme consists of a greedy policy improvement step followed by an m-step temporal difference (TD) policy evaluation step. We show that both the synchronous (entire state space evaluation) and asynchronous (a uniformly sampled set of substates) versions of the OPI scheme converge to the optimal value function and an optimal joint policy rollout.
  arXiv  Detail & Related papers  (2024-10-19T17:00:23Z)
• Projected Off-Policy Q-Learning (POP-QL) for Stabilizing Offline Reinforcement Learning [57.83919813698673]
  Projected Off-Policy Q-Learning (POP-QL) is a novel actor-critic algorithm that simultaneously reweights off-policy samples and constrains the policy to prevent divergence and reduce value-approximation error. In our experiments, POP-QL not only shows competitive performance on standard benchmarks, but also out-performs competing methods in tasks where the data-collection policy is significantly sub-optimal.
  arXiv  Detail & Related papers  (2023-11-25T00:30:58Z)
• Counterfactual Conservative Q Learning for Offline Multi-agent Reinforcement Learning [54.788422270960496]
  We propose a novel multi-agent offline RL algorithm, named CounterFactual Conservative Q-Learning (CFCQL) CFCQL calculates conservative regularization for each agent separately in a counterfactual way and then linearly combines them to realize an overall conservative value estimation. We prove that it still enjoys the underestimation property and the performance guarantee as those single agent conservative methods do, but the induced regularization and safe policy improvement bound are independent of the agent number.
  arXiv  Detail & Related papers  (2023-09-22T08:10:25Z)
• PAC: Assisted Value Factorisation with Counterfactual Predictions in Multi-Agent Reinforcement Learning [43.862956745961654]
  Multi-agent reinforcement learning (MARL) has witnessed significant progress with the development of value function factorization methods. In this paper, we show that in partially observable MARL problems, an agent's ordering over its own actions could impose concurrent constraints. We propose PAC, a new framework leveraging information generated from Counterfactual Predictions of optimal joint action selection.
  arXiv  Detail & Related papers  (2022-06-22T23:34:30Z)
• Sparse Feature Selection Makes Batch Reinforcement Learning More Sample Efficient [62.24615324523435]
  This paper provides a statistical analysis of high-dimensional batch Reinforcement Learning (RL) using sparse linear function approximation. When there is a large number of candidate features, our result sheds light on the fact that sparsity-aware methods can make batch RL more sample efficient.
  arXiv  Detail & Related papers  (2020-11-08T16:48:02Z)
• Kalman meets Bellman: Improving Policy Evaluation through Value Tracking [59.691919635037216]
  Policy evaluation is a key process in Reinforcement Learning (RL) We devise an optimization method, called Kalman Optimization for Value Approximation (KOVA) KOVA minimizes a regularized objective function that concerns both parameter and noisy return uncertainties.
  arXiv  Detail & Related papers  (2020-02-17T13:30:43Z)

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.