OGPO: Offline Goal-conditioned Policy Optimization for Recoverable Vision-Language-Action Models

Key points

• To address this limitation, we propose OGPO, an Offline Goal-conditioned Policy Optimization framework for recoverable VLA post-training.1
• Instead of treating offline demonstrations merely as step-wise action labels, OGPO adaptively relabels trajectories with semantic key states and constructs goal-conditioned process rewards to optimize the policy for reaching these task-progress states.1
• In this way, OGPO shifts VLA post-training from trajectory-level action imitation to key-state reachability learning, enabling policies to acquire a more robust ability to complete goal states from feasible off-demonstration states without requiring additional online interaction.1
• Moreover, evaluations on LIBERO-Plus and our proposed LIBERO-ReAct, a mid-execution perturbation benchmark, further demonstrate that OGPO achieves stronger robustness and goal-directed recoverability under distribution shifts and recoverable off-demonstration states.1

Sources (1)

• [1]OGPO: Offline Goal-conditioned Policy Optimization for Recoverable Vision-Language-Action Models
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 9, 03:04 AM

  Vision-language-action (VLA) models have shown strong promise in robotic manipulation, yet their post-training still predominantly relies on trajectory-level supervised imitation over robot demonstrations.

  To address this limitation, we propose OGPO, an Offline Goal-conditioned Policy Optimization framework for recoverable VLA post-training.

Extractive summary: sentences quoted from the sources.