Failing Forward: Adaptive Failure-Informed Learning for Vision-Language-Action Models
Adaptive Failure-Informed Learning (AFIL) improves Vision-Language-Action (VLA) models by using failure trajectories as adaptive negative guidance during policy sampling. The framework employs a Dual Action Generator architecture that jointly learns from both successful and failed behaviors. This allows the model to steer action generation away from failure modes, significantly boosting task success rates and robustness compared to standard success-only learning.