AI Paper Atlas

Research map / Video / Video-language understanding

An Empirical Study on How Video-LLMs Answer Video Questions

Chenhui Gou, Ziyu Ma, Zicheng Duan, Haoyu He, Feng Chen, Akide Liu, Bohan Zhuang, Jianfei Cai, Hamid Rezatofighi

CVPR 2026 · Video › Video-language understanding · approach: Video LLMs with fine-grained understanding · 4 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
block language-to-video attention in all layers (no visual info reaches text) (ablation)Video-LLMs · Video-MME / EgoSchemaperformance retained: retain at most 70% (LongVA, Video-MME), as little as 48% (LLaVA-Video, EgoSchema)Hurt
block language-to-video attention beyond about 60% of layers (ablation)Video-LLMs · Video-MME, MVBench, EgoSchemanormalized performanceNo clear effect
block language-to-video attention in layers between 50% and 60% depth (ablation)Video-LLMs · all benchmarksnormalized performanceHurt
knock out video temporal or spatial self-attention in all layers (ablation)Video-LLMs · Video-MME, MVBench, EgoSchemaaccuracyNo clear effect
knock out language-to-video attention in all layers (ablation)Video-LLMs · Video-MME, MVBench, EgoSchemaaccuracyHurt
knock out a window of 4 critical intermediate layers (e.g. layers 12-16 of 28) (ablation)Video-LLMs · VideoQA benchmarksnormalized performanceHurt
exit video tokens after layer 18LLaVA-Video-7B · Video-MMEaccuracy: 62.4 -> 62.0 (attention FLOPs 64.3%)No clear effect
exit video tokens after layer 18 and spatial-only attention in layers 1-8LLaVA-Video-7B · Video-MMEaccuracy: 62.4 -> 60.0 (attention FLOPs 37.6%)Hurt
exit video tokens after layer 18 and spatial-only attention in layers 1-8LLaVA-OneVision-7B · EgoSchemaaccuracy: 65.2 -> 65.2 (attention FLOPs 37.5%)No clear effect
exit video tokens after layer 18LLaVA-OneVision-7B · MVBenchaccuracy: 58.3 -> 57.3Hurt

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