Research

Paper

TESTING March 05, 2026

HiMAP-Travel: Hierarchical Multi-Agent Planning for Long-Horizon Constrained Travel

Authors

The Viet Bui, Wenjun Li, Yong Liu

Abstract

Sequential LLM agents fail on long-horizon planning with hard constraints like budgets and diversity requirements. As planning progresses and context grows, these agents drift from global constraints. We propose HiMAP-Travel, a hierarchical multi-agent framework that splits planning into strategic coordination and parallel day-level execution. A Coordinator allocates resources across days, while Day Executors plan independently in parallel. Three key mechanisms enable this: a transactional monitor enforcing budget and uniqueness constraints across parallel agents, a bargaining protocol allowing agents to reject infeasible sub-goals and trigger re-planning, and a single policy trained with GRPO that powers all agents through role conditioning. On TravelPlanner, HiMAP-Travel with Qwen3-8B achieves 52.78% validation and 52.65% test Final Pass Rate (FPR). In a controlled comparison with identical model, training, and tools, it outperforms the sequential DeepTravel baseline by +8.67~pp. It also surpasses ATLAS by +17.65~pp and MTP by +10.0~pp. On FlexTravelBench multi-turn scenarios, it achieves 44.34% (2-turn) and 37.42% (3-turn) FPR while reducing latency 2.5x through parallelization.

Metadata

arXiv ID: 2603.04750
Provider: ARXIV
Primary Category: cs.AI
Published: 2026-03-05
Fetched: 2026-03-06 14:20

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.04750v1</id>\n    <title>HiMAP-Travel: Hierarchical Multi-Agent Planning for Long-Horizon Constrained Travel</title>\n    <updated>2026-03-05T02:55:53Z</updated>\n    <link href='https://arxiv.org/abs/2603.04750v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.04750v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>Sequential LLM agents fail on long-horizon planning with hard constraints like budgets and diversity requirements. As planning progresses and context grows, these agents drift from global constraints. We propose HiMAP-Travel, a hierarchical multi-agent framework that splits planning into strategic coordination and parallel day-level execution. A Coordinator allocates resources across days, while Day Executors plan independently in parallel. Three key mechanisms enable this: a transactional monitor enforcing budget and uniqueness constraints across parallel agents, a bargaining protocol allowing agents to reject infeasible sub-goals and trigger re-planning, and a single policy trained with GRPO that powers all agents through role conditioning. On TravelPlanner, HiMAP-Travel with Qwen3-8B achieves 52.78% validation and 52.65% test Final Pass Rate (FPR). In a controlled comparison with identical model, training, and tools, it outperforms the sequential DeepTravel baseline by +8.67~pp. It also surpasses ATLAS by +17.65~pp and MTP by +10.0~pp. On FlexTravelBench multi-turn scenarios, it achieves 44.34% (2-turn) and 37.42% (3-turn) FPR while reducing latency 2.5x through parallelization.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.AI'/>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.CL'/>\n    <published>2026-03-05T02:55:53Z</published>\n    <arxiv:comment>33 pages, v1</arxiv:comment>\n    <arxiv:primary_category term='cs.AI'/>\n    <author>\n      <name>The Viet Bui</name>\n    </author>\n    <author>\n      <name>Wenjun Li</name>\n    </author>\n    <author>\n      <name>Yong Liu</name>\n    </author>\n  </entry>"
}