mlsanityData to 7 Oct 2026

Training the Orchestrator: A Supervised Approach to End-to-End PDDL Planning with LLM Agents

arXiv 2606.21740 ↗First posted 19 Jun 2026cs.AIcited by 0

On the map: LLM agents and tool use → General LLM agent frameworks and benchmarks

Abstract (arXiv)

Translating natural-language planning intent into verified plans is a longstanding challenge: people communicate goals in language, while classical planners require formal PDDL specifications. Recent agentic frameworks bridge this gap by orchestrating a pool of specialized repair agents inside a verifier-checked refinement loop, but the orchestrator at the centre is itself a prompted frontier LLM, paying a frontier-LLM API call at every refinement step. We present HALO (Hybrid Agent-Learned Orchestrator), which trains the orchestrator from refinement trajectories that an external verifier has certified as ending in valid plans, across 11 PDDL domains. HALO pairs a small QLoRA-tuned policy with three hardcoded rules for trivially decidable selections, and operates over an expanded 21-agent action space. Unlike approaches that prompt a frontier LLM at every step or learn an orchestrator from sparse end-of-episode rewards, our key observation is that the verifier already provides strong guidance: every accepted trajectory is a sequence of demonstrably correct (state, agent) decisions, directly usable as supervision. Across PlanBench, Natural Plan, and classical planning benchmarks, HALO matches or exceeds the GPT-5-mini prompted baseline on success rate, sits within three percentage points of the stronger Gemini-3-Flash prompted baseline, reduces orchestration cost by more than an order of magnitude (0.18 to 0.004 per task against GPT-5-mini, roughly 45× cheaper; roughly 15× cheaper than Gemini-3-Flash), and cuts total LLM calls per episode by 40 to 50 percent.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in General LLM agent frameworks and benchmarks, one of 16 directions in LLM agents and tool use. Builds on 43 of its 62 references that are in the map. First posted on arXiv 19 Jun 2026.

Builds on

43 of 62 references.

Closest ancestors · 6

Same direction, newest first

  1. 2025End-to-end PDDL Planning with Hardcoded and Dynamic Agents10 Dec
  2. 2025SynWorld: Virtual Scenario Synthesis for Agentic Action Knowledge Refinement4 Apr
  3. 2024Agent-FLAN: Designing Data and Methods of Effective Agent Tuning for Large Language Models19 Mar
  4. 2024KnowAgent: Knowledge-Augmented Planning for LLM-Based Agents5 Mar
  5. 2023AgentTuning: Enabling Generalized Agent Abilities for LLMs19 Oct
  6. 2023FireAct: Toward Language Agent Fine-tuning9 Oct

Other references · 37

  1. 2023Generative Agents: Interactive Simulacra of Human Behaviorcited by 3,500
  2. 2023Tree of Thoughts: Deliberate Problem Solving with Large Language Modelscited by 3,155
  3. 2023HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging Facecited by 1,173
  4. 2023Graph of Thoughts: Solving Elaborate Problems with Large Language Modelscited by 873
  5. 2023Reasoning with Language Model is Planning with World Modelcited by 846
  6. 2023LLM+P: Empowering Large Language Models with Optimal Planning Proficiencycited by 496
  7. 2024Training Language Models to Self-Correct via Reinforcement Learningcited by 405
  8. 2022PlanBench: An Extensible Benchmark for Evaluating Large Language Models on Planning and Reasoning about Changecited by 372
  9. 2024WebRL: Training LLM Web Agents via Self-Evolving Online Curriculum Reinforcement Learningcited by 243
  10. 2023Leveraging Pre-trained Large Language Models to Construct and Utilize World Models for Model-based Task Planningcited by 204
  11. 2024Trial and Error: Exploration-Based Trajectory Optimization for LLM Agentscited by 194
  12. 2023Synapse: Trajectory-as-Exemplar Prompting with Memory for Computer Controlcited by 180
25 more
  1. 2024LLMs Can't Plan, But Can Help Planning in LLM-Modulo Frameworkscited by 178
  2. 2024AlphaMath Almost Zero: Process Supervision without Processcited by 160
  3. 2023Generalized Planning in PDDL Domains with Pretrained Large Language Modelscited by 139
  4. 2025OpenCodeReasoning: Advancing Data Distillation for Competitive Codingcited by 106
  5. 2023ToolChain*: Efficient Action Space Navigation in Large Language Models with A* Searchcited by 103
  6. 2024Chain of Thoughtlessness? An Analysis of CoT in Planningcited by 100
  7. 2024Q*: Improving Multi-step Reasoning for LLMs with Deliberative Planningcited by 95
  8. 2023CLIN: A Continually Learning Language Agent for Rapid Task Adaptation and Generalizationcited by 88
  9. 2023Dynamic Planning with a LLMcited by 60
  10. 2023Coupling Large Language Models with Logic Programming for Robust and General Reasoning from Textcited by 59
  11. 2024TinyAgent: Function Calling at the Edgecited by 55
  12. 2024Can Language Models Serve as Text-Based World Simulators?cited by 50
  13. 2024Large Language Models Can Solve Real-World Planning Rigorously with Formal Verification Toolscited by 47
  14. 2024PlanRAG: A Plan-then-Retrieval Augmented Generation for Generative Large Language Models as Decision Makerscited by 42
  15. 2024OSCAR: Operating System Control via State-Aware Reasoning and Re-Planningcited by 37
  16. 2024Planetarium: A Rigorous Benchmark for Translating Text to Structured Planning Languagescited by 32
  17. 2024NL2Plan: Robust LLM-Driven Planning from Minimal Text Descriptionscited by 22
  18. 2024Unlocking the Future: Exploring Look-Ahead Planning Mechanistic Interpretability in Large Language Modelscited by 21
  19. 2025Classical Planning with LLM-Generated Heuristics: Challenging the State of the Art with Python Codecited by 18
  20. 2024TRIP-PAL: Travel Planning with Guarantees by Combining Large Language Models and Automated Plannerscited by 17
  21. 2024PROC2PDDL: Open-Domain Planning Representations from Textscited by 15
  22. 2024PDDLEGO: Iterative Planning in Textual Environmentscited by 14
  23. 2024UrbanLLM: Autonomous Urban Activity Planning and Management with Large Language Modelscited by 13
  24. 2025How Far Are LLMs from Symbolic Planners? An NLP-Based Perspectivecited by 6
  25. 2024Graph Neural Network Based Action Ranking for Planningcited by 3

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026PDDLCoder: Agentic PDDL Generation for LLM-Assisted Symbolic Planning
  2. 2025Learning When to Plan: Efficiently Allocating Test-Time Compute for LLM Agents
  3. 2023AdaPlanner: Adaptive Planning from Feedback with Language Models
  4. 2024Testing and Understanding Erroneous Planning in LLM Agents through Synthesized User Inputs
  5. 2026Agentic LLM Planning via Step-Wise PDDL Simulation: An Empirical Characterisation
  6. 2025Toward PDDL Planning Copilot
  7. 2026GRASP: Generating, Revising, and Assessing for Strategic Planning with Agentic AI
  8. 2026PIVOT: Bridging Planning and Execution in LLM Agents via Trajectory Refinement