arXiv cs.AI· Dongyin Zhuo, Fengjunjie Pan, Nenad Petrovic, Alois Knoll·· 12 小时前AI 评分24
Porimon：基于长短期知识增强生成（LSTKAG）的 LLM 宝可梦对战智能体
Porimon: An LLM-Based Pok\'emon Battle Agent Enhanced by Long/Short-Term Knowledge Augmented Generation
AI 导读
研究者提出长短期知识增强生成机制 LSTKAG，并据此设计 LLM 宝可梦对战智能体 Porimon，无需额外微调即可利用当前任务的历史状态并检索相似任务的经验摘要。在 15,000 场锦标赛式对战中，经超参数优化的 Porimon 显著优于此前研究提出的 PokéLLMon 和基于规则的启发式玩家。消融实验显示，扩展游戏信息检索的变体表现更好，但长期 KAG 的贡献尚无定论。
正文 · 原文
Abstract:In this paper, we use Pokémon Battles as a case study to investigate how to improve the performance of LLM-based agents in tasks that require opponent-aware planning without additional fine-tuning. We propose Long/Short-Term Knowledge Augmented Generation (LSTKAG), a mechanism that enables LLM-based agents to leverage past states of the current task and retrieve experience summaries from similar previous task instances based on the current state. Based on LSTKAG, we design Porimon, an LLM-based agent structure for Pokémon Battles. For optimization, we introduce an external API for precise damage calculation and more detailed information about the game. We conduct tournament-like evaluation experiments comprising 15,000 battles for hyperparameter optimization, ablation studies, and performance evaluation. The results indicate that Porimon-based players with hyperparameter optimization significantly outperform players based on PokéLLMon, an LLM-based agent structure proposed in previous research, and the rule-based heuristic player. Furthermore, our ablation study shows that Porimon variants outperform the one without extension in game information retrieval, which shows the contribution of that extension. However, the current experiment results are inconclusive regarding the contribution of Long-Term KAG. These results suggest that introducing external resources, information from previous states of the current task, and experience summaries from similar previous task instances could elevate the performance of LLM-based agents designed for tasks requiring opponent-aware planning.
| Comments: | 6 pages, 5 figures, accepted by FLLM 2026 |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2609.32544 [cs.AI] |
| (or arXiv:2609.32544v2 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2609.32544 arXiv-issued DOI via DataCite |
Submission history
From: Dongyin Zhuo [view email]
[v1]
Sat, 26 Sep 2026 12:22:58 UTC (376 KB)
[v2]
Thu, 15 Oct 2026 19:54:57 UTC (390 KB)
来源：arXiv cs.AI · arxiv.org