Hierarchical Expert Prompt for Large-Language-Model: An Approach Defeat Elite AI in TextStarCraft II for the First Time

2025-02-16

AI Summary

To address the failure of large language models (LLMs) in multi-level decision-making within TextStarCraft II—stemming from domain knowledge gaps and inadequate subtask prioritization—this paper proposes the Hierarchical Expert Prompting (HEP) framework. HEP structurally injects tactical domain knowledge and introduces an importance-aware task decomposition mechanism coupled with dynamic weighted scheduling, enabling precise modeling of complex real-time strategic behaviors. For the first time, an LLM equipped with HEP achieves stable victories over the highest-difficulty (Elite) built-in AI across all difficulty levels in TextStarCraft II benchmark evaluations, significantly outperforming existing baselines. The source code, prompt templates, and battle replays are fully open-sourced, establishing a reproducible and extensible paradigm for LLM-driven complex game AI.

Technology Category

• Humans and AI: Human-Aware Planning and Behavior Prediction
• Multiagent Systems: Adversarial Agents
• Search and Optimization: Heuristic Search

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information

Abstract

Since the emergence of the Large Language Model (LLM), LLM has been widely used in fields such as writing, translating, and searching. However, there is still great potential for LLM-based methods in handling complex tasks such as decision-making in the StarCraft II environment. To address problems such as lack of relevant knowledge and poor control over subtasks of varying importance, we propose a Hierarchical Expert Prompt (HEP) for LLM. Our method improves the understanding of game situations through expert-level tactical knowledge, improving the processing quality of tasks of varying importance through a hierarchical framework. Our approach defeated the highest level (Elite) standard built-in agent in TextStarCraft II for the first time and consistently outperformed the baseline method in other difficulties. Our experiments suggest that the proposed method is a practical solution for tackling complex decision-making challenges. The replay video can be viewed on https://www.bilibili.com/video/BV1uz42187EF and https://youtu.be/dO3PshWLV5M, and our codes have been open-sourced on https://github.com/luchang1113/HEP-LLM-play-StarCraftII.

Problem

• Enhance LLM decision-making in complex tasks
• Address knowledge gaps in StarCraft II
• Improve subtask control via hierarchical framework

Innovation

• Hierarchical Expert Prompt for LLM
• Defeated Elite AI in TextStarCraft II
• Improved decision-making through hierarchical framework