Chronos: 一种基于物理信息的全历史框架用于非马尔可夫长时域操作
Chronos: A Physics-Informed Full-History Framework for Non-Markovian Long-Horizon Manipulation
AI总结：
针对现有策略依赖马尔可夫假设导致记忆依赖任务失败的问题，提出Chronos框架，将观测历史作为策略潜状态，通过选择性状态空间模型传播因果历史状态，并利用二阶薛定谔桥预测加速度场，在16个模拟和4个真实任务中显著超越基线。
摘要：
通用机器人策略应建模为动力系统，但许多VLA和生成式模仿策略仍依赖当前观测或短时间窗口。这种马尔可夫捷径在记忆依赖的操作中失败：相同观测在不同历史后可能要求不同动作。我们提出Chronos，一个基于物理信息的全历史框架，用于非马尔可夫长时域操作。关键思想是将观测历史从辅助上下文提升为策略动力学的潜状态。在每个物理控制步，Chronos通过融合观测和本体感觉形成一个状态代表令牌，使得令牌序列与物理时间一一对齐。选择性状态空间模型传播这个因果历史状态，通过隐式最大似然估计（IMLE）条件化一个多模态粗动作先验。该先验随后由二阶薛定谔启发桥细化，预测加速度场，产生更平滑且更物理的机器人运动。在16个模拟任务和4个真实世界实验中，Chronos在精密插入、一般操作和记忆依赖的长时域控制上进行了评估。在RMBench上，成功需要记住任务阶段，Chronos达到73.6%的平均成功率，超过马尔可夫VLA基线pi0.5 62.4个百分点，相对增益6.6倍，同时使用少10倍的参数。它还超过记忆VLA Mem-0 22.8个百分点，同时使用少30倍以上的参数。在使用单个RGB相机的真实世界双臂实验中，Chronos在四个任务上达到78%的平均成功率，包括三个记忆依赖任务上的72%，而pi0.5整体为7%，记忆依赖子集为0%。这些结果表明，历史不应被视为辅助上下文，而应作为操作策略的潜状态。
英文摘要：
General-purpose robot policies should be modeled as dynamical systems, yet many VLA and generative imitation policies still rely on present observations or short windows. This Markovian shortcut fails in memory-dependent manipulation: identical observations can demand different actions after different histories. We present Chronos, a physics-informed full-history framework for non-Markovian long-horizon manipulation. The key idea is to elevate observation history from auxiliary context to the latent state of the policy dynamics. At each physical control step, Chronos forms one state-representative token by fusing observation and proprioception, so the token sequence is aligned one-to-one with physical time. A selective state space model propagates this causal historical state, which conditions a multimodal coarse action prior through implicit maximum likelihood estimation (IMLE). This prior is then refined by a second-order Schrodinger-inspired bridge that predicts acceleration fields, yielding smoother and more physically grounded robot motion. Across 16 simulated tasks and 4 real-world experiments, Chronos is evaluated on precision insertion, general manipulation, and memory-dependent long-horizon control. On RMBench, where success requires remembering task phase, Chronos achieves 73.6% average success, outperforming Markovian VLA baseline pi0.5 by +62.4 percentage points, a 6.6x relative gain, while using 10x fewer parameters. It also surpasses the memory VLA Mem-0 by 22.8 points while using over 30x fewer parameters. In real-world dual-arm experiments using a single RGB camera, Chronos achieves 78% average success over four tasks, including 72% on the three memory-dependent tasks, whereas pi0.5 achieves 7% overall and 0% on the memory-dependent subset. These results suggest that history should not be treated as auxiliary context, but as the latent state of the manipulation policy.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
