RUMAD: Reinforcement-Unifying Multi-Agent Debate

• URL: http://arxiv.org/abs/2602.23864v1
• Date: Fri, 27 Feb 2026 10:04:26 GMT
• Title: RUMAD: Reinforcement-Unifying Multi-Agent Debate
• Authors: Chao Wang, Han Lin, Huaze Tang, Huijing Lin, Wenbo Ding,
• Abstract summary: Multi-agent debate (MAD) systems leverage collective intelligence to enhance reasoning capabilities.<n>Existing approaches struggle to simultaneously optimize accuracy, consensus formation, and computational efficiency.<n>This work presents RUMAD, a novel framework that formulates dynamic communication topology control in MAD as a reinforcement learning (RL) problem.
• Score: 15.05837715937063
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Multi-agent debate (MAD) systems leverage collective intelligence to enhance reasoning capabilities, yet existing approaches struggle to simultaneously optimize accuracy, consensus formation, and computational efficiency. Static topology methods lack adaptability to task complexity variations, while external LLM-based coordination risks introducing privileged knowledge that compromises debate neutrality. This work presents RUMAD (Reinforcement-Unifying Multi-Agent Debate), a novel framework that formulates dynamic communication topology control in MAD as a reinforcement learning (RL) problem. RUMAD employs a content-agnostic observation scheme that captures high-level debate dynamics avoiding access to raw agent reasoning content. RUMAD uses a multi-objective reward to model solution quality, cohesion and efficiency. A PPO-trained controller dynamically adjusts edge weights in the communication graph, while a dual-threshold mechanism enables fine-grained control over both agent activation and information visibility. Experimental evaluation across MMLU, GSM8K, and GPQA benchmarks demonstrates that RUMAD achieves substantial efficiency gains, reducing token costs by over 80\%, while still improving reasoning accuracy compared to single LLM model and multiple MAD baselines. Notably, RUMAD trained exclusively on MMLU exhibits robust zero-shot generalization to out-of-domain (OOD) tasks, indicating that the learned communication strategies capture task-independent principles of effective multi-agent coordination. These results establish RUMAD as a efficient and robust approach for deploying multi-agent reasoning application with practical resource constraints.

Related papers

• Search-R2: Enhancing Search-Integrated Reasoning via Actor-Refiner Collaboration [49.9937230730202]
  We propose Search-R2, a novel Actor-Refiner collaboration framework that enhances reasoning through targeted intervention.<n>Our approach decomposes the generation process into an Actor, which produces initial reasoning trajectories.<n>We show that Search-R2 consistently outperforms strong RAG and RL-based baselines across model scales.
  arXiv  Detail & Related papers  (2026-02-03T15:32:09Z)
• RISER: Orchestrating Latent Reasoning Skills for Adaptive Activation Steering [62.63376387138257]
  We propose a plug-and-play intervention framework that adaptively steers large language models (LLMs) reasoning in activation space.<n>RISER constructs a library of reusable reasoning vectors and employs a lightweight Router to dynamically compose them for each input.<n>The Router is optimized via reinforcement learning under task-level rewards, activating latent cognitive primitives in an emergent and compositional manner.
  arXiv  Detail & Related papers  (2026-01-14T08:04:33Z)
• Multi-Agent Deep Reinforcement Learning for Collaborative UAV Relay Networks under Jamming Atatcks [36.380478794869234]
  This paper formulates this challenge as a cooperative Multi-Agent Reinforcement Learning (MARL) problem, solved using the Training with Decentralized Execution (CTDE) framework.<n>Our proposed framework significantly outperforms baselines, increasing total system throughput by approximately 50% while simultaneously achieving a near-zero collision rate.<n>A key finding is that the agents develop an emergent anti-jamming strategy without explicit programming.
  arXiv  Detail & Related papers  (2025-12-09T08:11:21Z)
• Agentic AI Reasoning for Mobile Edge General Intelligence: Fundamentals, Approaches, and Directions [74.35421055079655]
  Large language models (LLMs) have enabled an emergence of agentic artificial intelligence (AI) with powerful reasoning and autonomous decision-making capabilities.<n>Mobile Edge General Intelligence (MEGI) brings real-time, privacy-preserving reasoning to the network edge.<n>We propose a joint optimization framework for efficient LLM reasoning deployment in MEGI.
  arXiv  Detail & Related papers  (2025-09-27T10:53:48Z)
• Agentic Reinforced Policy Optimization [66.96989268893932]
  Large-scale reinforcement learning with verifiable rewards (RLVR) has demonstrated its effectiveness in harnessing the potential of large language models (LLMs) for single-turn reasoning tasks.<n>Current RL algorithms inadequately balance the models' intrinsic long-horizon reasoning capabilities and their proficiency in multi-turn tool interactions.<n>We propose Agentic Reinforced Policy Optimization (ARPO), a novel agentic RL algorithm tailored for training multi-turn LLM-based agents.
  arXiv  Detail & Related papers  (2025-07-26T07:53:11Z)
• Application of LLM Guided Reinforcement Learning in Formation Control with Collision Avoidance [1.1718316049475228]
  Multi-Agent Systems (MAS) excel at accomplishing complex objectives through the collaborative efforts of individual agents.<n>In this paper, we introduce a novel framework that aims to overcome the challenge of designing an effective reward function.<n>By giving large language models (LLMs) on the prioritization of tasks, our framework generates reward functions that can be dynamically adjusted online.
  arXiv  Detail & Related papers  (2025-07-22T09:26:00Z)
• MAD-Spear: A Conformity-Driven Prompt Injection Attack on Multi-Agent Debate Systems [4.931691794637798]
  Multi-agent debate (MAD) systems leverage collaborative interactions among large language models (LLMs) to improve reasoning capabilities.<n>We introduce MAD-Spear, a targeted prompt injection attack that compromises a small subset of agents but significantly disrupts the overall MAD process.
  arXiv  Detail & Related papers  (2025-07-17T12:09:39Z)
• Heterogeneous Group-Based Reinforcement Learning for LLM-based Multi-Agent Systems [25.882461853973897]
  We propose Multi-Agent Heterogeneous Group Policy Optimization (MHGPO), which guides policy updates by estimating relative reward advantages.<n>MHGPO eliminates the need for Critic networks, enhancing stability and reducing computational overhead.<n>We also introduce three group rollout sampling strategies that trade off between efficiency and effectiveness.
  arXiv  Detail & Related papers  (2025-06-03T10:17:19Z)
• Revisiting Multi-Agent Debate as Test-Time Scaling: A Systematic Study of Conditional Effectiveness [50.29739337771454]
  Multi-agent debate (MAD) approaches offer improved reasoning, robustness, and diverse perspectives over monolithic models.<n>This paper conceptualizes MAD as a test-time computational scaling technique, distinguished by collaborative refinement and diverse exploration capabilities.<n>We conduct a comprehensive empirical investigation comparing MAD with strong self-agent test-time scaling baselines on mathematical reasoning and safety-related tasks.
  arXiv  Detail & Related papers  (2025-05-29T01:02:55Z)
• ReMA: Learning to Meta-think for LLMs with Multi-Agent Reinforcement Learning [53.817538122688944]
  We introduce Reinforced Meta-thinking Agents (ReMA) to elicit meta-thinking behaviors from Reasoning of Large Language Models (LLMs)<n>ReMA decouples the reasoning process into two hierarchical agents: a high-level meta-thinking agent responsible for generating strategic oversight and plans, and a low-level reasoning agent for detailed executions.<n> Empirical results from single-turn experiments demonstrate that ReMA outperforms single-agent RL baselines on complex reasoning tasks.
  arXiv  Detail & Related papers  (2025-03-12T16:05:31Z)
• Effective Multi-Agent Deep Reinforcement Learning Control with Relative Entropy Regularization [6.441951360534903]
  Multi-Agent Continuous Dynamic Policy Gradient (MACDPP) was proposed to tackle the issues of limited capability and sample efficiency in various scenarios controlled by multiple agents. It alleviates the inconsistency of multiple agents' policy updates by introducing the relative entropy regularization to the Training with Decentralized Execution (CTDE) framework with the Actor-Critic (AC) structure.
  arXiv  Detail & Related papers  (2023-09-26T07:38:19Z)

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.