From Knowing to Doing Precisely: A General Self-Correction and Termination Framework for VLA models

TL;DR

This paper introduces VLA-SCT, a lightweight, training-free self-correction framework for vision-language-action models that improves task success and recognition of completion in embodied agents, enhancing robustness in complex environments.

Contribution

The paper presents a novel self-correcting, control loop framework for VLA models that operates without additional training to improve manipulation success and task termination detection.

Findings

Significant increase in grasping success rates.

Improved task completion recognition.

Enhanced robustness across datasets.

Abstract

While vision-language-action (VLA) models for embodied agents integrate perception, reasoning, and control, they remain constrained by two critical weaknesses: first, during grasping tasks, the action tokens generated by the language model often exhibit subtle spatial deviations from the target object, resulting in grasp failures; second, they lack the ability to reliably recognize task completion, which leads to redundant actions and frequent timeout errors. To address these challenges and enhance robustness, we propose a lightweight, training-free framework, VLA-SCT. This framework operates as a self-correcting control loop, combining data-driven action refinement with conditional logic for termination. Consequently, compared to baseline approaches, our method achieves consistent improvements across all datasets in the LIBERO benchmark, significantly increasing the success rate of fine manipulation tasks and ensuring accurate task completion, thereby promoting the deployment of more reliable VLA agents in complex, unstructured environments.