发表机构
University of California San Diego（加州大学圣迭戈分校）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
FACT是一种故障感知因果世界-动作模型，通过将错误动作转化为训练目标，在模拟与真实双臂操作任务中提升了世界-动作模型的性能，减少了成功偏差导致的未来幻觉。
AI 中文摘要
近期的世界-动作模型（World-Action Models，WAMs）研究表明，将策略与未来预测协同训练可为动作生成提供物理先验。依托视频模型的未来预测能力，许多WAMs会生成未来视频并通过逆动力学模型恢复动作，或将这些预测视频作为动作生成的目标条件。但在这两种场景下，世界模型大多仅在成功演示数据上训练，几乎没有动力预测错误动作的后果。我们提出FACT，一种因果世界-动作模型，它基于执行的动作预测未来视频与任务进展。这种基于动作的接口允许故障 rollout（滚动）监督动作后果，将错误动作转化为有效未来目标而非被丢弃。故障感知训练使进展预测器同时感知成功与失败的动作结果，推理时可选择性用于对采样动作候选进行评分。在模拟及真实世界双臂操作任务上的大量实验表明，FACT优于多个现有基线，随训练中纳入故障数据而性能提升，且减少了错误动作下受成功偏差影响的未来幻觉。
英文摘要
Recent world-action models (WAMs) show that co-training policies with future prediction can provide physical priors for action generation. Building on the future-prediction ability of video models, many WAMs generate future videos and recover actions with inverse-dynamics models, or use these predicted videos as goal conditions for action generation. In both cases, the world model is trained mostly on successful demonstrations and has little reason to predict the consequences of bad actions. We introduce FACT, a causal World-Action Model that predicts future video and task progress conditioned on the executed action. This action-conditioned interface allows failure rollouts to supervise action consequences, turning bad actions into valid future targets rather than being discarded. Failure-aware training makes the progress predictor aware of both successful and failed action outcomes, which can optionally be used to score sampled action candidates at inference. Extensive experiments on simulation and real-world bimanual manipulation tasks show that FACT outperforms many existing baselines, improves as failure data are incorporated into training, and reduces success-biased future hallucination under bad actions. See more details at https://fact-wam.github.io/
