HyPOLE: Hyperproperty-Guided Multi-Agent Reinforcement Learning under Partial Observation
Summary
HyPOLE introduces a framework for multi-agent reinforcement learning under partial observability that uses hyperproperty-guided learning via HyperLTL temporal logic, integrated with centralized training for decentralized execution, and demonstrates improvements over baselines on SMAC, MessySMAC, and WildFire benchmarks.
View Cached Full Text
Cached at: 07/01/26, 05:36 AM
# HyPOLE: Hyperproperty-Guided Multi-Agent Reinforcement Learning under Partial Observation Source: [https://arxiv.org/abs/2606.30966](https://arxiv.org/abs/2606.30966) [View PDF](https://arxiv.org/pdf/2606.30966) > Abstract:Formal specification is a powerful tool to guide the learning process and provides significant advantages over reward shaping: \(1\) mathematical rigor; \(2\) expressiveness to specify objectives and constraints, and \(3\) the ability to define tactics to achieve objectives\. However, these benefits remain largely unexplored in the context of Multi\-Agent Reinforcement Learning \(MARL\)\. This paper introduces HyPOLE, a novel framework for MARL under partial observability, where learning is guided by the expressive power of the so\-called hyperproperties and, in particular, the temporal logic HyperLTL\. We integrate Centralized Training for Decentralized Execution \(CTDE\) techniques with HyPOLE to synthesize decentralized policies, and our evaluation on SMAC, MessySMAC, and WildFire benchmark demonstrates clear advantages over baselines\. ## Submission history From: Arshia Rafieioskouei \[[view email](https://arxiv.org/show-email/71a408e0/2606.30966)\] **\[v1\]**Mon, 29 Jun 2026 23:03:05 UTC \(14,134 KB\)
Similar Articles
Hierarchical Experimentalist Agents
Introduces HExA, a training-free framework enabling LLMs to learn through active experimentation and skill reuse, achieving up to 77% success on the new Interphyre physics benchmark, a large improvement over existing agents.
LambdaPO: A Lambda Style Policy Optimization for Reasoning Language Models
Introduces LambdaPO, a novel reinforcement learning framework that improves upon GRPO by decomposing advantage estimation into pairwise preference comparisons and adding a semantic density reward, achieving better performance on math reasoning tasks.
Training Observable Control Policies to Expose Agent State Through Actions
This paper investigates using reinforcement learning to train observable control policies that enable estimation of an agent's state from its actions, with applications in multiagent coordination and monitoring under communications constraints.
Learning policy representations in multiagent systems
OpenAI researchers propose a general framework for learning representations of agent policies in multiagent systems using minimal interaction data, casting the problem as representation learning with applications to competitive control and cooperative communication environments.
Learning and Reusing Policy Decompositions for Hierarchical Generalized Planning with LLM Agents
This paper introduces HCL-GP, a dynamic policy-learning framework that integrates generalized planning and hierarchical task decomposition to enable LLM-based agents to learn and reuse executable policy components, significantly improving performance on the AppWorld benchmark.