[论文] From Single-Actor Environments to Multi-Actor Event Worlds: Can LLMs B...

论文概要

研究领域: ML 作者: Juntong Ni, Zewen Liu, Bohan Wang, Lige Zhang, Wei Jin 发布时间: 2026-10-24 arXiv: 2610.30488

中文摘要

世界模型帮助智能体在行动前评估后果，但许多真实系统涉及多个行动者：他们各自决策、即使智能体不行动也会演化、且智能体只能通过事件日志部分观测。本文把这类系统形式化为「事件世界」，追问：LLM 能否从日志与智能体提议的动作预测未来的观测？作者用人类 Diplomacy 对局与真实 Polymarket 交易构建互补基准，层级框架先比较有无他人动作时的预测，再测多步 rollout、对从未采取过的动作的预测、以及对智能体的实际效用。微调和更多数据在给出所有记录动作时改善预测；但只给智能体自身动作时，当其他行动者强烈影响结果，模型表现挣扎，更大的模型与更长的历史帮助有限。这一差距也限制了多步预测与智能体决策——模型高估计划成功率。不过，想象经验与世界模型初始化仍能改善策略学习。预测其他行动者的动作，仍是事件世界建模的核心挑战。

原文摘要

World models help agents assess the consequences of their actions before acting. Many real systems, however, involve multiple actors who make their own decisions, can evolve even when the agent takes no action, and are only partially observed through event logs. We formalize these systems as event worlds and ask whether large language models can predict future observations from the log and the agent's proposed actions. We build complementary benchmarks from human Diplomacy games and real Polymarket trades. Our hierarchical framework compares prediction with and without other actors' actions, then tests multi-step rollouts, predictions for actions never taken, and usefulness to agents. Fine-tuning and more data improve prediction when all actors' recorded actions are given. With only the agent's action, models struggle when other actors strongly affect the outcome, and larger models and longer histories help little. This gap also limits multi-step prediction and agent decisions, with models overestimating plan success. Yet imagined experience and initialization from world models can still improve policy learning. Anticipating other actors' actions remains a central challenge for event world modeling.

---

*自动采集于 2026-10-24*

#论文 #arXiv #ML #小凯