Dejavu: Post-Deployment Learning for Embodied Agents via Experience Feedback
This addresses the limitation of fixed policies in embodied agents, enabling continual refinement after deployment, though it is incremental as it builds on existing vision-language-action policies.
The paper tackles the problem of embodied agents being unable to learn new knowledge after deployment by proposing Dejavu, a post-deployment learning framework that uses an Experience Feedback Network to retrieve and condition on past successful actions, resulting in significant improvements in adaptability, robustness, and success rates over frozen baselines.
Embodied agents face a fundamental limitation: once deployed in real-world environments to perform specific tasks, they are unable to acquire new useful knowledge to enhance task performance. In this paper, we propose a general post-deployment learning framework called Dejavu, which employs an Experience Feedback Network (EFN) and augments the frozen Vision-Language-Action (VLA) policy with retrieved execution memories. EFN automatically identifies contextually successful prior action experiences and conditions action prediction on this retrieved guidance. We adopt reinforcement learning with semantic similarity rewards on EFN to ensure that the predicted actions align with past successful behaviors under current observations. During deployment, EFN continually enriches its memory with new trajectories, enabling the agent to exhibit "learning from experience" despite fixed weights. Experiments across diverse embodied tasks show that EFN significantly improves adaptability, robustness, and success rates over frozen baselines. These results highlight a promising path toward embodied agents that continually refine their behavior after deployment.