AI Paper Atlas

Research map / Video / Video-language understanding

AdaSpark: Adaptive Sparsity for Efficient Long-Video Understanding

Handong Li, Zikang Liu, Longteng Guo, Tongtian Yue, Yepeng Tang, Xinxin Zhu, Chuanyang Zheng, Ziming Wang, Zhibin Wang, Jun Song, Cheng Yu, Bo Zheng and 1 more

CVPR 2026 · Video › Video-language understanding · approach: Long video understanding · 0 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 AdaS-Attn to the dense model (ablation)Qwen2.5-VL-3B · TFLOPsTFLOPs: 299.5 -> 213.1Improved
add AdaS-FFN to the dense model (ablation)Qwen2.5-VL-3B · TFLOPsTFLOPs: 299.5 -> 216.9Improved
combine AdaS-Attn and AdaS-FFN (ablation)Qwen2.5-VL-3B · TFLOPsTFLOPs: 299.5 -> 128.5Improved
add AdaS-FFN only (ablation)Qwen2.5-VL-3B · MLVUM-avg: 65.2 -> 63.9Hurt
add AdaS-FFN only (ablation)Qwen2.5-VL-3B · CharadesSTAmIoU: 43.6 -> 41.7Hurt
add AdaS-FFN only (ablation)Qwen2.5-VL-3B · VideoMMEaccuracy: 60.3 -> 62.8Improved
add AdaS-Attn only (ablation)Qwen2.5-VL-3B · MLVUM-avg: 65.2 -> 66.0Improved
add mean compensation to AdaS-FFN (ablation)AdaSpark 3B · MLVUM-avg: 56.4 -> 57.5Improved
add mean compensation to AdaS-FFN (ablation)AdaSpark 3B · VideoMMEaccuracy: 54.1 -> 55.7Improved
add mean compensation to AdaS-FFN (ablation)AdaSpark 3B · CharadesSTAmIoU: 35.0 -> 36.1Improved
use dynamic Top-P instead of static Top-K (ablation)AdaSpark 3B · VideoMMEaccuracy: 52.6 -> 55.7Improved
use Top-P instead of uniform sampling (ablation)AdaSpark 3B · VideoMMEaccuracy: 50.2 -> 55.7Improved

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