VLA models from VLM adaptation can be pruned 12-30% via multi-module joint scheme based on divergence signals while keeping ~90% performance on LIBERO without post-pruning recovery, unlike standard criteria that collapse.
hub Canonical reference
arXiv preprint arXiv:2510.24795 (2025)
Canonical reference. 80% of citing Pith papers cite this work as background.
14
Pith papers citing it
Background
80% of classified citations
hub tools
citation-role summary
background 4
baseline 1
citation-polarity summary
years
2026 14representative citing papers
VLA language backbones show high redundancy on manipulation benchmarks, with half the LLM blocks removable and even two blocks sufficient to recover baseline performance after fine-tuning, unlike vision and action pathways.
A3 reframes dynamic action chunk commitment in VLA models as self-speculative prefix verification, accepting the longest continuous sequence of actions that satisfies consensus-ordered conditional invariance and prefix-closed sequential consistency.
CF-VLA uses a coarse initialization over endpoint velocity followed by single-step refinement to achieve strong performance with low inference steps on CALVIN, LIBERO, and real-robot tasks.
VLA models exhibit a compute-bound VLM phase followed by a memory-bound action phase on edge hardware; DP-Cache and V-AEFusion reduce redundancy and enable pipeline parallelism for up to 6x speedup on NPUs with marginal task degradation.
vla.cpp is a unified C++ runtime that serves multiple VLA architectures with flow-matching and diffusion patterns, matching SOTA performance on LIBERO while running on low-memory embedded hardware.
BPF prunes embodied LLM controllers iteratively during RL (and optionally SFT) to achieve superior size-performance-throughput trade-offs compared to post-training pruning or smaller dense models on the RobotxR1 autonomous driving pipeline.
M²-VLA shows that generalized VLMs can serve as direct backbones for robotic manipulation by selectively extracting task-critical features via Mixture of Layers and adding Meta Skill Modules for efficient trajectory learning.
SV-VLA uses infrequent heavy VLA planning of action chunks plus a lightweight closed-loop verifier to achieve both efficiency and robustness in dynamic robot control.
FASTER adds a Horizon-Aware Schedule to flow VLAs that compresses immediate-action denoising to one step while keeping long-horizon trajectory quality, lowering real-robot reaction latency.
ElegantVLA accelerates VLA models up to 3.77x by dynamically scheduling compute across vision, language, and action components without retraining the base model.
SpanVLA reduces action generation latency via flow-matching conditioned on history and improves robustness by training on negative-recovery samples with GRPO and a dedicated reasoning dataset.
A literature survey that unifies fragmented work on attacks, defenses, evaluations, and deployment challenges for Vision-Language-Action models in robotics.
Threading optimization of RTAC for VLA models reduces end-to-end latency and improves stability on low-cost agricultural robotic arms without changing the policy.
citing papers explorer
-
vla.cpp: A Unified Inference Runtime for Vision-Language-Action Models
vla.cpp is a unified C++ runtime that serves multiple VLA architectures with flow-matching and diffusion patterns, matching SOTA performance on LIBERO while running on low-memory embedded hardware.