arXiv Computer Vision August 27, 2026 By Jintao Cheng, Weibin Li, Haozhe Wang, Gang Wang, Yipu Zhang, Xiaoyu Tang, Jin Wu, Xieyuanli Chen, Yunhui Liu, Wei Zhang

Training-Free Interaction-Aligned Visual Token Pruning for Efficient Embodied Manipulation

Read the original on arXiv Computer Vision →

The paper introduces Interaction‑Aligned Pruning (IAprune), a training‑free method for visual token pruning in embodied manipulation tasks. IAprune jointly decides per‑frame budget and token selection, using semantic‑motion spatial agreement to choose between conservative and aggressive coverage, and applies geometric residual correction to focus on under‑represented boundaries. Experiments on four policies, three simulation benchmarks, and a real‑robot platform show that IAprune matches unpruned performance on LIBERO while achieving up to 1.54× speed‑up and 1.48× acceleration on a real robot.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Computer Vision.

arXiv AI

Jun 3

See Less, Specify More: Visual Evidence Budgets for Generalizable VLAs

arXiv:2606. 02735v1 Announce Type: cross Abstract: Generalization remains a central bottleneck for vision-language-action (VLA) models: under distractors, appearance shifts, and semantically similar tasks, the policy must often infer local execution details from coarse instructions while also deciding which parts of the image matter for control.

By Yueh-Hua Wu, Tatsuya Matsushima, Kei Ota

llms robotics multimodal

arXiv AI

Jun 30

SA-VLA: State-aware tokenizer for improving Vision-Language-Action Models' performance

arXiv:2606. 30113v1 Announce Type: cross Abstract: Discrete action tokenization provides a compact interface for autoregressive VLA policies, but accurately recovering continuous robot actions from discrete codes remains challenging.

By Tengyue Jiang, Chunpu Xu, Jiayue Kang, Yao Mu

llms nlp robotics multimodal

arXiv Machine Learning

Jul 21

Foresight Residual RL for Long-Horizon Robot Manipulation with Vision-Language-Action Models

arXiv:2607. 16506v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) policies offer strong general-purpose manipulation priors, but often fail on tight-tolerance, contact-rich assembly due to long-horizon credit assignment and subtask coupling: a state that is geometrically successful for the current skill can be brittle for downstream skills.

By Yuhan Liu, Xinyu Zhang, Litao Liu, Abdeslam Boularias

reinforcement-learning robotics multimodal

arXiv AI

4d ago

Pointing-VLA: Typed Spatial Grounding Interfaces for Vision-Language-Action Manipulation

arXiv:2608.23138v1 Announce Type: cross Abstract: Vision-language-action (VLA) models often expose spatial grounding through autoregressive text coordinates or opaque action tokens, creating brittle...

By Xiwen Chen, Zelin Li, Zhiruo Zhou, Huiming Chen, Chenwei Wang, Xiaojun Zhu

agents robotics fine-tuning multimodal benchmarks

arXiv Machine Learning

Jul 21

Patch Policy: Efficient Embodied Control via Dense Visual Representations

arXiv:2607. 18236v1 Announce Type: cross Abstract: Pretrained dense visual features from Vision Transformers (ViTs) are powerful yet have been underutilized in robot learning.

By Gaoyue Zhou, Zichen Jeff Cui, Ada Langford, Bowen Tan, Yann LeCun, Lerrel Pinto

llms robotics fine-tuning efficiency multimodal benchmarks

arXiv AI

Jun 30

Fast Enough to Act: Spatio-Temporal Visual Token Merging for Low-Latency Robotic VLMs and VLAs

arXiv:2606. 29350v1 Announce Type: cross Abstract: Vision-language models and vision-language action models endow the robot with unprecedented capabilities.

By Junzhou Chen, Jindong Wang, Gang Zhou

llms nlp robotics efficiency multimodal

One email a morning, machine-written

Related stories

See Less, Specify More: Visual Evidence Budgets for Generalizable VLAs

SA-VLA: State-aware tokenizer for improving Vision-Language-Action Models' performance

Foresight Residual RL for Long-Horizon Robot Manipulation with Vision-Language-Action Models

Pointing-VLA: Typed Spatial Grounding Interfaces for Vision-Language-Action Manipulation

Patch Policy: Efficient Embodied Control via Dense Visual Representations

Fast Enough to Act: Spatio-Temporal Visual Token Merging for Low-Latency Robotic VLMs and VLAs