Planner Matters! An Efficient and Unbalanced Multi-agent Collaboration Framework for Long-horizon Planning

• URL: http://arxiv.org/abs/2605.02168v1
• Date: Mon, 04 May 2026 02:58:05 GMT
• Title: Planner Matters! An Efficient and Unbalanced Multi-agent Collaboration Framework for Long-horizon Planning
• Abstract summary: Language model (LM)-based agents have demonstrated promising capabilities in automating complex tasks from natural language instructions.<n>We propose an enhanced multi-agent framework that decomposes automation into three roles: a planner for high-level decision-making, an actor for task execution, and a memory manager for contextual reasoning.
• Score: 32.56971325545821
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Language model (LM)-based agents have demonstrated promising capabilities in automating complex tasks from natural language instructions, yet they continue to struggle with long-horizon planning and reasoning. To address this, we propose an enhanced multi-agent framework that decomposes automation into three roles: a planner for high-level decision-making, an actor for task execution, and a memory manager for contextual reasoning. While this modular decomposition aligns with established design patterns, our core contribution lies in a systematic compute-allocation analysis, revealing that planning is the dominant factor influencing task performance. Execution and memory management require significantly less compute and model capacity to achieve competitive results. Building on these insights, we introduce a planner-centric reinforcement learning approach, which exclusively optimizes the planner using trajectory-level rewards from a VLM-as-judge, while freezing the other components. Extensive experiments on benchmarks spanning web navigation, OS control, and tool use demonstrate that concentrating model capacity and learning on high-level planning yields robust and compute-efficient improvements in long-horizon agent automation. Our code is publicly released.

Related papers

• EASy: Towards Efficient LLM-Based Agentic System [54.10819421625103]
  We propose EASy, a trainable agentic framework that jointly optimize task performance and computational efficiency through reinforcement learning.<n> EASy equips an LLM-based orchestrator with explicit knowledge of the capability and cost profiles of heterogeneous executors, enabling context-sensitive coordination beyond performance-only routing.<n>It further introduces a milestone-plan-act workflow that decomposes complex tasks into manageable milestones, constructs dependency-aware execution graphs, assigns suitable executors, and parallelizes independent steps while adapting subsequent decisions to intermediate outcomes.
  arXiv  Detail & Related papers  (2026-08-05T08:50:36Z)
• PlanningBench: Generating Scalable and Verifiable Planning Data for Evaluating and Training Large Language Models [52.48858778580074]
  Planning is a fundamental capability for large language models (LLMs)<n>PlanningBench is a framework for generating scalable, diverse, verifiable planning data for both evaluation and training.
  arXiv  Detail & Related papers  (2026-05-20T08:10:15Z)
• From Coarse to Fine: Self-Adaptive Hierarchical Planning for LLM Agents [24.79854509473032]
  Large language model-based agents have emerged as powerful approaches for solving dynamic and multi-step tasks.<n>We propose textbfAdaPlan-H, a self-adaptive hierarchical planning mechanism that mimics human planning strategies.<n>Our method significantly improves task execution success rates while mitigating overplanning at the planning level.
  arXiv  Detail & Related papers  (2026-04-25T07:54:23Z)
• RoboAgent: Chaining Basic Capabilities for Embodied Task Planning [46.248451288196435]
  This paper focuses on embodied task planning, where an agent acquires visual observations from the environment and executes atomic actions to accomplish a given task.<n>We propose RoboAgent, a capability-driven planning pipeline in which the model actively invokes different sub-capabilities.<n>We use a multi-stage paradigm that consists of: (1) behavior cloning with expert plans, (2) DAgger training using trajectories collected by the model, and (3) reinforcement learning guided by an expert policy.
  arXiv  Detail & Related papers  (2026-04-09T04:01:27Z)
• Scale-Plan: Scalable Language-Enabled Task Planning for Heterogeneous Multi-Robot Teams [11.964765256346857]
  Long-horizon task planning for heterogeneous multi-robot systems is essential for deploying collaborative teams in real-world environments.<n>We present Scale-Plan, a scalable LLM-assisted framework that generates compact, task-relevant problem representations from natural language instructions.<n>By filtering irrelevant information prior to planning, Scale-Plan enables efficient decomposition, allocation, and long-horizon plan generation.
  arXiv  Detail & Related papers  (2026-03-09T18:13:18Z)
• MagicAgent: Towards Generalized Agent Planning [73.21129030631421]
  We present textbfMagicAgent, a series of foundation models specifically designed for generalized agent planning.<n>We introduce a lightweight and scalable synthetic data framework that generates high-quality trajectories across diverse planning tasks.<n>We show that MagicAgent-32B and MagicAgent-30B-A3B achieve superior performance across diverse open-source benchmarks.
  arXiv  Detail & Related papers  (2026-02-22T01:39:16Z)
• TodoEvolve: Learning to Architect Agent Planning Systems [68.48983335970901]
  TodoEvolve is a meta-planning paradigm that autonomously synthesizes and dynamically revises task-specific planning.<n>PlanFactory provides a common interface for heterogeneous planning patterns.<n>TodoEvolve consistently surpasses carefully engineered planning modules while maintaining economical API costs and runtime overhead.
  arXiv  Detail & Related papers  (2026-02-08T06:37:01Z)
• ParaCook: On Time-Efficient Planning for Multi-Agent Systems [62.471032881396496]
  Large Language Models (LLMs) exhibit strong reasoning abilities for planning long-horizon, real-world tasks.<n>We present ParaCook, a benchmark for time-efficient collaborative planning.
  arXiv  Detail & Related papers  (2025-10-13T16:47:07Z)
• Learning When to Plan: Efficiently Allocating Test-Time Compute for LLM Agents [35.79575378215309]
  Training large language models (LLMs) to reason via reinforcement learning (RL) significantly improves their problem-solving capabilities.<n>We introduce a conceptual framework formalizing dynamic planning for LLM agents, enabling them to flexibly decide when to allocate test-time compute for planning.<n>Experiments on the Crafter environment show that dynamic planning agents trained with this approach are more sample-efficient and consistently achieve more complex objectives.
  arXiv  Detail & Related papers  (2025-09-03T18:00:13Z)
• Unlocking Reasoning Potential in Large Langauge Models by Scaling Code-form Planning [94.76546523689113]
  We introduce CodePlan, a framework that generates and follows textcode-form plans -- pseudocode that outlines high-level, structured reasoning processes. CodePlan effectively captures the rich semantics and control flows inherent to sophisticated reasoning tasks. It achieves a 25.1% relative improvement compared with directly generating responses.
  arXiv  Detail & Related papers  (2024-09-19T04:13:58Z)
• Improving Planning with Large Language Models: A Modular Agentic Architecture [6.394504490512812]
  Large language models (LLMs) often struggle with tasks that require multi-step reasoning or goal-directed planning.<n>We propose an agentic architecture, the Modular Agentic Planner (MAP), in which planning is accomplished via the recurrent interaction of specialized modules.<n>We find that MAP yields significant improvements over both standard LLM methods.
  arXiv  Detail & Related papers  (2023-09-30T00:10:14Z)
• Hierarchical Imitation Learning with Vector Quantized Models [77.67190661002691]
  We propose to use reinforcement learning to identify subgoals in expert trajectories. We build a vector-quantized generative model for the identified subgoals to perform subgoal-level planning. In experiments, the algorithm excels at solving complex, long-horizon decision-making problems outperforming state-of-the-art.
  arXiv  Detail & Related papers  (2023-01-30T15:04:39Z)

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.