Multi-Agent Deep Research: Training Multi-Agent Systems with M-GRPO

• URL: http://arxiv.org/abs/2511.13288v2
• Date: Tue, 18 Nov 2025 03:13:18 GMT
• Title: Multi-Agent Deep Research: Training Multi-Agent Systems with M-GRPO
• Authors: Haoyang Hong, Jiajun Yin, Yuan Wang, Jingnan Liu, Zhe Chen, Ailing Yu, Ji Li, Zhiling Ye, Hansong Xiao, Yefei Chen, Hualei Zhou, Yun Yue, Minghui Yang, Chunxiao Guo, Junwei Liu, Peng Wei, Jinjie Gu,
• Abstract summary: Current training methods train a unified large language model for all agents in the system.<n>This may limit the performances due to different underlying distributions for different agents.<n>We propose M-GRPO, a hierarchical extension of Group Relative Policy Optimization for vertical Multi-agent systems.
• Score: 24.532870400949424
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Multi-agent systems perform well on general reasoning tasks. However, the lack of training in specialized areas hinders their accuracy. Current training methods train a unified large language model (LLM) for all agents in the system. This may limit the performances due to different distributions underlying for different agents. Therefore, training multi-agent systems with distinct LLMs should be the next step to solve. However, this approach introduces optimization challenges. For example, agents operate at different frequencies, rollouts involve varying sub-agent invocations, and agents are often deployed across separate servers, disrupting end-to-end gradient flow. To address these issues, we propose M-GRPO, a hierarchical extension of Group Relative Policy Optimization designed for vertical Multi-agent systems with a main agent (planner) and multiple sub-agents (multi-turn tool executors). M-GRPO computes group-relative advantages for both main and sub-agents, maintaining hierarchical credit assignment. It also introduces a trajectory-alignment scheme that generates fixed-size batches despite variable sub-agent invocations. We deploy a decoupled training pipeline in which agents run on separate servers and exchange minimal statistics via a shared store. This enables scalable training without cross-server backpropagation. In experiments on real-world benchmarks (e.g., GAIA, XBench-DeepSearch, and WebWalkerQA), M-GRPO consistently outperforms both single-agent GRPO and multi-agent GRPO with frozen sub-agents, demonstrating improved stability and sample efficiency. These results show that aligning heterogeneous trajectories and decoupling optimization across specialized agents enhances tool-augmented reasoning tasks.

Related papers

• Dr. MAS: Stable Reinforcement Learning for Multi-Agent LLM Systems [20.971694319263353]
  We propose Dr. MAS, a simple and stable RL training recipe for multi-agent LLM systems.<n>Dr. MAS uses agent-wise remedy: normalizing advantages per agent using each agent's own reward statistics.<n>Dr. MAS provides an end-to-end RL training framework for multi-agent LLM systems.
  arXiv  Detail & Related papers  (2026-02-09T16:13:39Z)
• MARTI-MARS$^2$: Scaling Multi-Agent Self-Search via Reinforcement Learning for Code Generation [64.2621682259008]
  Multi-Agent Reinforced Training and Inference Framework with Self-Search Scaling (MARTI-MARS2)<n>We propose a Multi-Agent Reinforced Training and Inference Framework with Self-Search Scaling (MARTI-MARS2) to integrate policy learning with multi-agent tree search.<n>We show that MARTI-MARS2 achieves 77.7%, outperforming strong baselines like GPT-5.1 on challenging code generation benchmarks.
  arXiv  Detail & Related papers  (2026-02-08T07:28:44Z)
• AgentArk: Distilling Multi-Agent Intelligence into a Single LLM Agent [57.10083973844841]
  AgentArk is a novel framework to distill multi-agent dynamics into the weights of a single model.<n>We investigate three hierarchical distillation strategies across various models, tasks, scaling, and scenarios.<n>By shifting the burden of computation from inference to training, the distilled models preserve the efficiency of one agent while exhibiting strong reasoning and self-correction performance of multiple agents.
  arXiv  Detail & Related papers  (2026-02-03T19:18:28Z)
• BOAD: Discovering Hierarchical Software Engineering Agents via Bandit Optimization [41.08366028094234]
  Large language models (LLMs) struggle to generalize to real-world software engineering problems.<n>Existing systems often rely on a single agent to handle the entire workflow-interpreting issues.<n>Motivated by how human engineers decompose complex problems, we propose structuring SWE agents as orchestrators coordinating specialized sub-agents.
  arXiv  Detail & Related papers  (2025-12-29T17:41:11Z)
• Multi-Agent Tool-Integrated Policy Optimization [67.12841355267678]
  Large language models (LLMs) increasingly rely on multi-turn tool-integrated planning for knowledge-intensive and complex reasoning tasks.<n>Existing implementations typically rely on a single agent, but they suffer from limited context length and noisy tool responses.<n>No existing methods support effective reinforcement learning post-training of tool-integrated multi-agent frameworks.
  arXiv  Detail & Related papers  (2025-10-06T10:44:04Z)
• AgentRL: Scaling Agentic Reinforcement Learning with a Multi-Turn, Multi-Task Framework [76.96794548655292]
  Large language models (LLMs) have sparked growing interest in building generalist agents that can learn through online interactions.<n>Applying reinforcement learning (RL) to train LLM agents in multi-turn, multi-task settings remains challenging due to lack of scalable infrastructure and stable training algorithms.<n>We present the AgentRL framework for scalable multi-turn, multi-task agentic RL training.
  arXiv  Detail & Related papers  (2025-10-05T13:40:01Z)
• InfiAgent: Self-Evolving Pyramid Agent Framework for Infinite Scenarios [28.65914611521654]
  InfiAgent is a Pyramid-like DAG-based Multi-Agent Framework that can be applied to textbfinfinite scenarios.<n>InfiAgent achieves 9.9% higher performance compared to ADAS (similar auto-generated agent framework)
  arXiv  Detail & Related papers  (2025-09-26T15:44:09Z)
• How to Train a Leader: Hierarchical Reasoning in Multi-Agent LLMs [16.853362180877593]
  We introduce a hierarchical multi-agent framework that trains only a single leader LLM to coordinate a team of untrained peer agents.<n>Our results highlight the effectiveness and efficiency of training a single, flexible leader for collaborative reasoning in multi-agent LLM systems.
  arXiv  Detail & Related papers  (2025-07-11T18:34:07Z)
• MALT: Improving Reasoning with Multi-Agent LLM Training [67.76186488361685]
  MALT (Multi-Agent LLM Training) is a novel post-training strategy that divides the reasoning process into generation, verification, and refinement steps.<n>On MATH, GSM8K, and CSQA, MALT surpasses the same baseline LLM with a relative improvement of 15.66%, 7.42%, and 9.40% respectively.
  arXiv  Detail & Related papers  (2024-12-02T19:30:36Z)
• EvoAgent: Towards Automatic Multi-Agent Generation via Evolutionary Algorithms [55.77492625524141]
  EvoAgent is a generic method to automatically extend specialized agents to multi-agent systems.<n>We show that EvoAgent can significantly enhance the task-solving capability of LLM-based agents.
  arXiv  Detail & Related papers  (2024-06-20T11:49:23Z)

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.