aiqbench analysis
This research addresses the critical deployment bottleneck of Vision-Language-Action models in edge robotics by demonstrating that standard pruning causes severe safety and performance degradation. The authors introduce GLUESTICK, a post-pruning recovery technique that restores near-dense model capabilities without retraining, enabling significant memory savings and speedups while maintaining safety constraints. This work advances the practical viability of deploying large-scale VLA systems on resource-constrained field robots.
Don't Run with Scissors: Pruning Breaks VLA Models but They Can Be Recovered
arXiv:2510.08464v2 Announce Type: replace-cross Abstract: Vision-Language-Action (VLA) models are a promising paradigm for field robotics, but their deployment is bottlenecked by strict edge SWaP-C constraints. Compression techniques such as quantization, distillation, and pruning are well-established for LLMs and VLMs. Pruning, however, remains particularly underexplored for VLA models. We surprisingly find that pruning VLA models yields drastic performance degradation and increased safety violations. Through spectral analysis, we identify structural properties of VLA weights that help explain this fragility. Based on this insight, we introduce GLUESTICK, a post-pruning recovery method that restores much of the original model's capability while preserving sparsity benefits. GLUESTICK computes a one-time corrective term by interpolating between the dense and pruned models in weight space. At inference, each pruned layer applies this correction to recover lost performance with minimal overhead. Our approach requires no additional training and is agnostic to the pruning algorithm. Across diverse VLA architectures and domains, GLUESTICK recovers up to 100% of the success rate and restores safety to near-dense levels, while saving up to 5 GB of memory and achieving up to 1.75x speedup. Additional material can be found at: https://gluestick-vla.github.io/
Abstract:Vision-Language-Action (VLA) models are a promising paradigm for field robotics, but their deployment is bottlenecked by strict edge SWaP-C constraints. Compression techniques such as quantization, distillation, and pruning are well-established for LLMs and VLMs. Pruning, however, remains particularly underexplored for VLA models. We surprisingly find that pruning VLA models yields drastic performance degradation and increased safety violations. Through spectral analysis, we identify structural properties of VLA weights that help explain this fragility. Based on this insight, we introduce GLUESTICK, a post-pruning recovery method that restores much of the original model's capability while preserving sparsity benefits. GLUESTICK computes a one-time corrective term by interpolating between the dense and pruned models in weight space. At inference, each pruned layer applies this correction to recover lost performance with minimal overhead. Our approach requires no additional training and is agnostic to the pruning algorithm. Across diverse VLA architectures and domains, GLUESTICK recovers up to 100% of the success rate and restores safety to near-dense levels, while saving up to 5 GB of memory and achieving up to 1.75x speedup. Additional material can be found at: this https URL
Robotics (cs.RO); Machine Learning (cs.LG)
arXiv:2510.08464 [cs.RO]
(or arXiv:2510.08464v2 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2510.08464 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 278 words
Related coverage
Large Models for Small Devices: Recent Advances and Empirical Analysis of Edge AI Deployment
arXiv · Aug 19, 2026
Compression Trinity: Exploring Sparsity, Quantization, and Low-Rank Approximations for LLM Compression
arXiv · Aug 27, 2026
Breaking the Compression Barrier: Cross-Architecture Compression Boundary Learning via Reverse Regrowth
arXiv · Aug 19, 2026
When Fewer Layers Break More Chains: Layer Pruning Harms Test-Time Scaling in LLMs
arXiv · Jul 21, 2026