AI Paper Atlas

Research map / Video / Video-language understanding

Video-LevelGauge: Investigating Contextual Positional Bias in Video Language Models.

Hou Xia, Zheren Fu, Fangcan Ling, Jiajun Li, Yi Tu, Zhendong Mao, Yongdong Zhang

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
use a commercial model instead of an open-source modelLVLMs · Video-LevelGauge MCQApositional bias (P_ran, P_var): Gemini 2.5 Pro P_ran 2.0, P_var 0.9 vs Qwen2.5-VL-7B 12.4, 11.7Improved
enable multimodal reasoning mode (ablation)GLM-4.5V 108B · Video-LevelGauge MCQAP_ran: 3.4 -> 2.7Improved
enable multimodal reasoning mode (ablation)MiMo-VL-RL 7B · Video-LevelGauge MCQAP_var: 4.9 -> 1.8Improved
increase context length (ablation)LVLMs (8 models) · Video-LevelGaugepositional biasHurt
use more complex contexts (long video, multi-video, interleaved video-text, long text) instead of template or short video (ablation)LVLMs · Video-LevelGaugepositional biasHurt
scale up model size within the same series (ablation)InternVL3, LLaVA-Video, Qwen2.5-VL, LLaVA-OV · Video-LevelGauge MCQAP_var: LLaVA-Video 7B 3.5 -> 72B 1.2Improved
scale up Qwen2.5-VL from 7B to 32B (ablation)Qwen2.5-VL · Video-LevelGauge open-endedP_var: 13.8 -> 7.2Improved
increase probe length from 2 to 8 frames (ablation)InternVL3-8B, LLaVA-OV-7B, Qwen2.5-VL-7B · Video-LevelGauge MCQAS_meta: InternVL3-8B 59.2 -> 74.9Improved
increase probe length from 2 to 8 frames (ablation)InternVL3-8B, LLaVA-OV-7B, Qwen2.5-VL-7B · Video-LevelGauge MCQApositional bias (P_var): InternVL3-8B 5.8 -> 5.4No clear effect
use open-ended descriptive questions instead of MCQA (ablation)LVLMs · Video-LevelGaugepositional biasHurt
long video training (LongVA, LongVILA)LongVA 7B, LongVILA-1M 7B · Video-LevelGauge MCQAP_ran: LongVA 9.2, LongVILA-1M 11.5No clear effect
use a two-stage caption-then-LLM pipelineCaption + GPT-4 · Video-LevelGauge MCQAP_ran: 7.4 vs Qwen2.5-VL-7B 12.4Improved

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