Adaptive Collaboration with Humans: Metacognitive Policy Optimization for Multi-Agent LLMs with Continual Learning

• URL: http://arxiv.org/abs/2603.07972v1
• Date: Mon, 09 Mar 2026 05:18:07 GMT
• Title: Adaptive Collaboration with Humans: Metacognitive Policy Optimization for Multi-Agent LLMs with Continual Learning
• Abstract summary: We propose the Human-In-the-Loop Multi-Agent Collaboration (HILA) framework.<n>HILA trains agents to learn a metacognitive policy that governs when to solve problems autonomously and when to defer to a human expert.<n>Experiments on challenging mathematical and problem-solving benchmarks show that HILA, equipped with Dual-Loop Policy Optimization, consistently outperforms advanced MAS.
• Score: 12.114998959919978
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: While scaling individual Large Language Models (LLMs) has delivered remarkable progress, the next frontier lies in scaling collaboration through multi-agent systems (MAS). However, purely autonomous MAS remain ''closed-world'' systems, constrained by the static knowledge horizon of pre-trained models. This limitation makes them brittle on tasks requiring knowledge beyond training data, often leading to collective failure under novel challenges. To address this, we propose the Human-In-the-Loop Multi-Agent Collaboration (HILA) framework, a principled paradigm for human--agent collaboration. HILA trains agents to learn a metacognitive policy that governs when to solve problems autonomously and when to defer to a human expert. To operationalize this policy, we introduce Dual-Loop Policy Optimization, which disentangles immediate decision-making from long-term capability growth. The inner loop applies Group Relative Policy Optimization (GRPO) with a cost-aware reward to optimize deferral decisions, while the outer loop implements continual learning, transforming expert feedback into high-quality supervised signals that strengthen the agent's reasoning ability. Experiments on challenging mathematical and problem-solving benchmarks show that HILA, equipped with Dual-Loop Policy Optimization, consistently outperforms advanced MAS, establishing a principled foundation for collaborative and continually improving agentic systems.

Related papers

• Self-evolving LLM agents with in-distribution Optimization [60.05867547965365]
  Large Language Models (LLMs) have recently emerged as powerful controllers for interactive agents in complex environments.<n>We propose Q-Evolve, a self-evolving framework for LLM agents that unifies automatic process-reward labeling and policy learning.<n>We evaluate our method on AlfWorld, WebShop, and ScienceWorld, showing Q-Evolve outperforms strong baselines in sample efficiency, robustness, and overall task performance.
  arXiv  Detail & Related papers  (2026-06-05T15:09:52Z)
• AMATA: Adaptive Multi-Agent Trajectory Alignment for Knowledge-Intensive Question Answering [59.145044878490815]
  AMATA is an Adaptive Multi-Agent Trajectory Alignment framework.<n>It dynamically integrates external knowledge to improve response interpretability and factual grounding.<n>We introduce two principal innovations: (1) Intra-Trajectory Preference Learning, which learns objective-oriented preferences to prioritize critical agents, and (2) Inter-Agent Dependency Learning, which captures cross-agent tool dependencies.
  arXiv  Detail & Related papers  (2026-05-17T09:45:24Z)
• SelfAI: Building a Self-Training AI System with LLM Agents [79.10991818561907]
  SelfAI is a general multi-agent platform that combines a User Agent for translating high-level research objectives into standardized experimental configurations.<n>An Experiment Manager orchestrates parallel, fault-tolerant training across heterogeneous hardware while maintaining a structured knowledge base for continuous feedback.<n>Across regression, computer vision, scientific computing, medical imaging, and drug discovery benchmarks, SelfAI consistently achieves strong performance and reduces redundant trials.
  arXiv  Detail & Related papers  (2025-11-29T09:18:39Z)
• Maestro: Learning to Collaborate via Conditional Listwise Policy Optimization for Multi-Agent LLMs [23.590034731179824]
  We present Through Role Orchestration (Maestro), a principled paradigm for collaboration that structurally decouples cognitive modes.<n>Maestro uses a collective of parallel Execution Agents for diverse exploration and a specialized Central Agent for convergent, evaluative synthesis.<n>Experiments on mathematical reasoning and general problem-solving benchmarks demonstrate that Maestro, coupled with CLPO, consistently outperforms existing state-of-the-art multi-agent approaches.
  arXiv  Detail & Related papers  (2025-11-08T21:01:27Z)
• Strategic Coordination for Evolving Multi-agent Systems: A Hierarchical Reinforcement and Collective Learning Approach [0.0]
  Reinforcement learning offers a way to model sequential decision-making.<n>Agents take high-level strategies using MARL to group possible plans for action space reduction.<n>Low-level collective learning layer ensures efficient and decentralized coordinated decisions.
  arXiv  Detail & Related papers  (2025-09-22T17:58:45Z)
• Improving Retrospective Language Agents via Joint Policy Gradient Optimization [57.35348425288859]
  RetroAct is a framework that jointly optimize both task-planning and self-reflective evolution capabilities in language agents.<n>We develop a two-stage joint optimization process that integrates imitation learning and reinforcement learning.<n>We conduct extensive experiments across various testing environments, demonstrating RetroAct has substantial improvements in task performance and decision-making processes.
  arXiv  Detail & Related papers  (2025-03-03T12:54:54Z)
• From Novice to Expert: LLM Agent Policy Optimization via Step-wise Reinforcement Learning [62.54484062185869]
  We introduce StepAgent, which utilizes step-wise reward to optimize the agent's reinforcement learning process.<n>We propose implicit-reward and inverse reinforcement learning techniques to facilitate agent reflection and policy adjustment.
  arXiv  Detail & Related papers  (2024-11-06T10:35:11Z)
• Learning Reward and Policy Jointly from Demonstration and Preference Improves Alignment [58.049113055986375]
  We develop a single stage approach named Alignment with Integrated Human Feedback (AIHF) to train reward models and the policy.<n>The proposed approach admits a suite of efficient algorithms, which can easily reduce to, and leverage, popular alignment algorithms.<n>We demonstrate the efficiency of the proposed solutions with extensive experiments involving alignment problems in LLMs and robotic control problems in MuJoCo.
  arXiv  Detail & Related papers  (2024-06-11T01:20:53Z)
• ROMA-iQSS: An Objective Alignment Approach via State-Based Value Learning and ROund-Robin Multi-Agent Scheduling [44.276285521929424]
  We introduce a decentralized state-based value learning algorithm that enables agents to independently discover optimal states. Our theoretical analysis shows that our approach leads decentralized agents to an optimal collective policy. Empirical experiments further demonstrate that our method outperforms existing decentralized state-based and action-based value learning strategies.
  arXiv  Detail & Related papers  (2024-04-05T09:39:47Z)
• Large Language Model-based Human-Agent Collaboration for Complex Task Solving [94.3914058341565]
  We introduce the problem of Large Language Models (LLMs)-based human-agent collaboration for complex task-solving. We propose a Reinforcement Learning-based Human-Agent Collaboration method, ReHAC. This approach includes a policy model designed to determine the most opportune stages for human intervention within the task-solving process.
  arXiv  Detail & Related papers  (2024-02-20T11:03:36Z)
• 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)

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.