Zeva: In-Context Causal Learning for Generalizable Embodied Manipulation
2026-08-31 • Robotics
Robotics
AI summaryⓘ
The authors propose Zeva, a system that helps robots learn from their own actions in the real world without changing their main decision-making program. Zeva remembers the cause-and-effect of the robot's past actions and uses this knowledge to improve future decisions by adding context. Tests show that Zeva helps robots get better over time during actual tasks and that what the robot learns can be useful for different tasks as well. This approach allows robots to adapt on the spot without expensive retraining.
embodied manipulationin-context learningcausal interactionpolicy modelrobot deploymentcausal memoryself-evolutionvisual-language agentsreal-world roboticsgeneralization
Authors
Fu Chen, Xin Ding, Bingjia Huang, Xiangyu Li, Mingju Wang, Jiawei He, Kun Li, Wei Sun, Yunxin Liu, Hao Wu, Ting Cao
Abstract
Generalizable embodied manipulation remains difficult to achieve through pretraining alone, due to unseen physical conditions in the real world. We argue that robots need to learn from their own physical interactions on the fly during real-world deployment and use this knowledge to inform subsequent actions. We present Zeva, the first framework that enables in-context learning from a robot's own physical interaction experience while keeping the policy model frozen. Zeva employs a Causal Interaction Extractor to encode an executed action and its induced state change into a causal interaction signal, which is stored in a dual-timescale causal memory. For subsequent actions, relevant causal interaction signals are retrieved from memory and injected into the frozen policy model as context. Experiments in simulation and real-world manipulation demonstrate that Zeva achieves the best performance among the compared frontier VLAs and WAMs and, more importantly, enables self-evolution during deployment without gradient updates. Its success rate continues to improve as the robot accumulates interaction experience. Furthermore, the acquired interaction experience can generalize across tasks.