AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

SAFE-Pruner: Semantic Attention–Guided Future-Aware Token Pruning for Efficient Vision-Language-Action Manipulation

Shilin Ma, Chubin Zhang, Changyuan Wang, Yuji Wang, Yue Wu, Zixuan Wang, Jingqi Tian, Zheng Zhu, Yansong Tang

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Attention-based visual token pruning · 6 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add future-layer saliency forecast (ablation)SAFE-Pruner on OpenVLA-OFT · LIBERO (4 suites avg)success rate: 94.5 -> 96.4Improved
add future-layer saliency forecast (ablation)SAFE-Pruner on OpenVLA-OFT · LIBERO (4 suites avg)FLOPs (T): 2.141 -> 1.722Improved
adaptive key timestep selection instead of fixed interval (ablation)SAFE-Pruner on OpenVLA-OFT · LIBERO (4 suites avg)success rate: 95.8 -> 96.4Improved
adaptive key timestep selection instead of fixed interval (ablation)SAFE-Pruner on OpenVLA-OFT · LIBERO (4 suites avg)latency (ms): 51.98 -> 37.22Improved
SAFE-Pruner vs best prior pruningOpenVLA-OFT · LIBERO (4 suites avg)success rate: 94.5 (FastV) -> 96.4Improved
SAFE-Pruner vs FastV on real robotpi0.5 Astribot S1 · real robot avgsuccess rate: 69.0 -> 79.3Improved
SAFE-Pruner vs unpruned modelOpenVLA-OFT · LIBERO (4 suites avg)success rate: 96.8 -> 96.4Hurt
random token pruning (ablation)CogACT · SIMPLER Visual Matchingsuccess rate: 74.8 -> 20.9Hurt
combine SAFE-Pruner with HiF8 quantization (ablation)OpenVLA-OFT · LIBERO (4 suites avg)GPU memory (GB): 18.31 -> 11.03Improved
SAFE-Pruner on SIMPLER Variant Aggregation vs VLA-CacheCogACT · SIMPLER Variant Aggregationsuccess rate: 62.3 (VLA-Cache) vs 61.9No clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy