让AI学会在宝可梦红版中完成早期任务，避免陷入循环和误操作。
PokeRL: Reinforcement Learning for Pokemon Red
- 设计带地图遮蔽的环境包装，识别并阻止动作循环。
- 采用分层密集奖励机制，成功让智能体通过前两座道馆挑战。
- 专治常见失败模式，适合研究实用RL系统的开发者参考。
《宝可梦红》是一款长周期、稀疏奖励、部分可观测且操控机制奇特的角色扮演游戏，是强化学习的高难度基准。尽管近期工作表明，使用强化奖励塑造与工程化观测的PPO智能体可通关前两个道馆，但训练过程仍不稳定，常出现动作循环、菜单重复点击或无意义徘徊等现象。本文提出PokeRL，一个模块化系统，用于训练深度强化学习智能体完成《宝可梦红》早期任务，包括离开玩家家、探索帕拉特镇进入草丛区域，以及击败首个对手。主要贡献包括：基于PyBoy模拟器的循环感知环境包装器（含地图遮蔽）、多层防循环与防菜单重复机制，以及密集分层奖励设计。我们认为，像PokeRL这样显式建模失败模式的实用系统，是通往完整宝可梦联赛冠军智能体的必要中间步骤。代码已开源。
原文摘要 · Abstract (English)
Pokemon Red is a long-horizon JRPG with sparse rewards, partial observability, and quirky control mechanics that make it a challenging benchmark for reinforcement learning. While recent work has shown that PPO agents can clear the first two gyms using heavy reward shaping and engineered observations, training remains brittle in practice, with agents often degenerating into action loops, menu spam, or unproductive wandering. In this paper, we present PokeRL, a modular system that trains deep reinforcement learning agents to complete early game tasks in Pokemon Red, including exiting the player's house, exploring Pallet Town to reach tall grass, and winning the first rival battle. Our main contributions are a loop-aware environment wrapper around the PyBoy emulator with map masking, a multi-layer anti-loop and anti-spam mechanism, and a dense hierarchical reward design. We argue that practical systems like PokeRL, which explicitly model failure modes such as loops and spam, are a necessary intermediate step between toy benchmarks and full Pokemon League champion agents. Code is available at https://github.com/reddheeraj/PokemonRL
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。