让世界模型在测试时自适应调整，提升规划成功率。
AdaJEPA: An Adaptive Latent World Model
- 测试时通过闭环控制动态更新模型，无需额外数据。
- 仅需每步一次梯度更新，即可显著提高任务成功率。
- 适合需要实时适应环境变化的机器人规划场景。
潜在世界模型通过在紧凑的潜在空间中预测未来状态，实现从高维观测中进行规划。然而，这些模型通常在测试时保持冻结：当预测不准确时，规划可能失败，尤其是在测试分布发生偏移的情况下。为解决此问题，我们提出AdaJEPA，一种可在模型预测控制（MPC）闭环中进行测试时自适应的潜在世界模型。训练完成后，AdaJEPA执行首个动作块，利用观察到的下一状态转移作为自监督适应信号，并用更新后的模型重新规划。这一闭环更新过程持续校准世界模型，无需额外专家示范。在多种目标达成任务中，AdaJEPA仅需每轮MPC重规划一步梯度更新，即显著提升规划成功率。
原文摘要 · Abstract (English)
Latent world models enable planning from high-dimensional observations by predicting future states in a compact latent space. However, these models are typically kept frozen at test time: when their predictions become inaccurate, planning can fail, especially under test-time distribution shift. To address this, we propose AdaJEPA, an adaptive latent world model that performs test-time adaptation within the closed loop of model predictive control (MPC). After training, AdaJEPA plans and executes the first action chunk, uses the observed next-state transition as a self-supervised adaptation signal, and replans with the updated model. This closed-loop update continuously recalibrates the world model without additional expert demonstrations. Across a range of goal-reaching tasks, AdaJEPA substantially improves planning success with as few as one gradient step per MPC replanning step.
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。