AI Paper Atlas

Research map / Video / Video-language understanding

Divid: Disentangled Spatial-Temporal Modeling within LLMs for Temporally Grounded Video Understanding

Yepeng Tang, Weining Wang, Longteng Guo, Tongtian Yue, Wenxuan Wang, Chunjie Zhang, Jing Liu

ICLR 2026 · Video › Video-language understanding · approach: Video LLMs with fine-grained understanding

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 Top-K temporal-attention keyframe selection instead of uniform sampling (ablation)Divid · Charades-STAmIoU: 46.27 -> 47.33Improved
add Top-K temporal-attention keyframe selection instead of uniform sampling (ablation)Divid · ReXTimemIoU: 27.61 -> 29.71Improved
add token-wise soft-router instead of plain addition (ablation)Divid · Charades-STAmIoU: 46.81 -> 47.33Improved
add token-wise soft-router instead of plain addition (ablation)Divid · ReXTimemIoU: 29.11 -> 29.71Improved
disentangle spatial and temporal modeling inside the LLM instead of Slow-Fast concatenation (ablation)Divid · Charades-STAmIoU: 46.71 -> 47.33Improved
disentangle spatial and temporal modeling inside the LLM instead of Slow-Fast concatenation (ablation)Divid · ReXTimemIoU: 29.49 -> 29.71Improved
use disentangled Divid instead of full high-resolution frames (ablation)Divid · Charades-STATFLOPs: 28.2 -> 10.5 TFLOPsImproved
train on TempGCap instead of Momentor-10M (ablation)Divid · Charades-STAmIoU: 46.42 -> 47.33Improved
train on TempGCap instead of Momentor-1M (ablation)Divid · Charades-STAmIoU: 44.90 -> 47.33Improved
fuse at every decoder layer instead of only at output level (ablation)Divid · Charades-STAmIoU: 42.51 -> 47.33Improved
use separate LoRA attention parameters per branch instead of shared (ablation)Divid · Charades-STA/ReXTimemIoU: 47.33 -> 47.38 (Charades-STA mIoU); 29.71 -> 29.24 (ReXTime mIoU)No clear effect
increase temporal frames from 128 to 256 (ablation)Divid · ReXTimemIoU: 29.71 -> 30.35No 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