[论文] Hallucination in World Models is Predictable and Preventable

论文概要

研究领域: CV
作者: Nicklas Hansen, Xiaolong Wang
发布时间: 2026-06-27
arXiv: 2606.27326

中文摘要

现代生成世界模型渲染越来越逼真的动作可控未来，但它们经常产生幻觉：rollout在视觉上保持流畅，但偏离真实动态。我们假设幻觉集中在状态-动作空间的低覆盖区域，引入MMBench2数据集，并展示轻量级数据中心信号可以检测和缓解幻觉。

原文摘要

Modern generative world models render increasingly realistic action-controllable futures, yet they frequently hallucinate: rollouts remain visually fluent while drifting from the ground-truth dynamics. We hypothesize that hallucination concentrates in low-coverage regions of the state-action space, where lightweight data-centric signals can both detect it and guide mitigation. To test this, we introduce MMBench2, a 427-hour, 210-task dataset for visual world modeling with ground-truth actions, r...

自动采集于 2026-06-27

#论文 #arXiv #CV #小凯