Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2510.15259 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911544201183232 |
|---|---|
| author | Tang, Chenwei Long, Lin Liu, Xinyu Xing, Jingyu Wang, Zizhou Zhou, Joey Tianyi Du, Jiawei Zhen, Liangli Lv, Jiancheng |
| author_facet | Tang, Chenwei Long, Lin Liu, Xinyu Xing, Jingyu Wang, Zizhou Zhou, Joey Tianyi Du, Jiawei Zhen, Liangli Lv, Jiancheng |
| contents | Most commodity software lacks accessible Application Programming Interfaces (APIs), requiring autonomous agents to interact solely through pixel-based Graphical User Interfaces (GUIs). In this API-free setting, large language model (LLM)-based agents face severe efficiency bottlenecks: limited to local visual experiences, they make myopic decisions and rely on inefficient trial-and-error, hindering both skill acquisition and long-horizon planning. To overcome these limitations, we propose SAG-Agent, an experience-driven learning framework that structures an agent's raw pixel-level interactions into a persistent State-Action Graph (SAG). SAG-Agent mitigates inefficient exploration by topologically linking functionally similar but visually distinct GUI states, constructing a rich neighborhood of experience that enables the agent to generalize from a diverse set of historical strategies. To facilitate long-horizon reasoning, we design a novel hybrid intrinsic reward mechanism based on the graph topology, combining a state-value reward for exploiting known high-value pathways with a novelty reward that encourages targeted exploration. This approach decouples strategic planning from pure discovery, allowing the agent to effectively value setup actions with delayed gratification. We evaluate SAG-Agent in two complex, open-ended GUI-based decision-making environments (Civilization V and Slay the Spire), demonstrating significant improvements in exploration efficiency and strategic depth over the state-of-the-art methods. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_15259 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | SAG-Agent: Enabling Long-Horizon Reasoning in Strategy Games via Dynamic Knowledge Graphs Tang, Chenwei Long, Lin Liu, Xinyu Xing, Jingyu Wang, Zizhou Zhou, Joey Tianyi Du, Jiawei Zhen, Liangli Lv, Jiancheng Artificial Intelligence Most commodity software lacks accessible Application Programming Interfaces (APIs), requiring autonomous agents to interact solely through pixel-based Graphical User Interfaces (GUIs). In this API-free setting, large language model (LLM)-based agents face severe efficiency bottlenecks: limited to local visual experiences, they make myopic decisions and rely on inefficient trial-and-error, hindering both skill acquisition and long-horizon planning. To overcome these limitations, we propose SAG-Agent, an experience-driven learning framework that structures an agent's raw pixel-level interactions into a persistent State-Action Graph (SAG). SAG-Agent mitigates inefficient exploration by topologically linking functionally similar but visually distinct GUI states, constructing a rich neighborhood of experience that enables the agent to generalize from a diverse set of historical strategies. To facilitate long-horizon reasoning, we design a novel hybrid intrinsic reward mechanism based on the graph topology, combining a state-value reward for exploiting known high-value pathways with a novelty reward that encourages targeted exploration. This approach decouples strategic planning from pure discovery, allowing the agent to effectively value setup actions with delayed gratification. We evaluate SAG-Agent in two complex, open-ended GUI-based decision-making environments (Civilization V and Slay the Spire), demonstrating significant improvements in exploration efficiency and strategic depth over the state-of-the-art methods. |
| title | SAG-Agent: Enabling Long-Horizon Reasoning in Strategy Games via Dynamic Knowledge Graphs |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2510.15259 |