Don't Run with Scissors: Pruning Breaks VLA Models but They Can Be Recovered
This addresses the challenge of deploying large VLA models on resource-limited robotic hardware, offering a practical solution to mitigate safety and performance risks from pruning, though it is incremental as it builds on existing pruning techniques.
The paper tackles the problem of pruning Vision-Language-Action (VLA) models for robotic deployment, which causes severe performance degradation and safety issues, and introduces GLUESTICK, a post-pruning recovery method that restores much of the original functionality while maintaining sparsity, achieving competitive memory efficiency and improved success rates across tasks.
Vision-Language-Action (VLA) models have advanced robotic capabilities but remain challenging to deploy on resource-limited hardware. Pruning has enabled efficient compression of large language models (LLMs), yet it is largely understudied in robotics. Surprisingly, we observe that pruning VLA models leads to drastic degradation and increased safety violations. We introduce GLUESTICK, a post-pruning recovery method that restores much of the original model's functionality while retaining sparsity benefits. Our method performs a one-time interpolation between the dense and pruned models in weight-space to compute a corrective term. This correction is used during inference by each pruned layer to recover lost capabilities with minimal overhead. GLUESTICK requires no additional training, is agnostic to the pruning algorithm, and introduces a single hyperparameter that controls the tradeoff between efficiency and accuracy. Across diverse VLA architectures and tasks in manipulation and navigation, GLUESTICK achieves competitive memory efficiency while substantially recovering success rates and reducing safety violations. Additional material can be found at: https://gluestick-vla.github.io/.