AI Paper Atlas

Research map / Video / Video-language understanding

Native Active Perception as Reasoning for Omni-Modal Understanding

Zhenghao Xing, Ruiyang Xu, Yuxuan Wang, Jinzheng He, Ziyang Ma, Qize Yang, Yunfei Chu, Jin Xu, Junyang Lin, Chi-Wing Fu, Pheng-Ann Heng

ICML 2026 · Video › Video-language understanding · approach: Event and anomaly reasoning in video · 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
fine-tune on static QA pairs (standard SFT) instead of agentic trajectories (ablation)Qwen2.5-Omni-7B · LVBenchaccuracy: 43.0 -> 41.6Hurt
use Agentic SFT (OTA trajectories) instead of standard SFT (ablation)Qwen2.5-Omni-7B · LVBenchaccuracy: 41.6 -> 48.7Improved
add Agentic SFT to base model (ablation)Qwen2.5-Omni-7B · DailyOmniaccuracy: 60.1 -> 63.3Improved
add vanilla GRPO RL after Agentic SFT (ablation)OmniAgent · DailyOmniaccuracy: 63.3 -> 62.2Hurt
add vanilla GRPO RL after Agentic SFT (ablation)OmniAgent · MLVUaccuracy: 69.9 -> 69.9No clear effect
add vanilla GRPO RL after Agentic SFT (ablation)OmniAgent · LVBenchaccuracy: 48.7 -> 49.8Improved
use TAURA turn-level entropy rescaled advantage instead of vanilla GRPO (ablation)OmniAgent · MLVUaccuracy: 69.9 -> 71.1Improved
use TAURA instead of vanilla GRPO (ablation)OmniAgent · DailyOmniaccuracy: 62.2 -> 64.8Improved
use TAURA instead of vanilla GRPO (ablation)OmniAgent · LVBenchaccuracy: 49.8 -> 50.5Improved
increase maximum turn limit K from 6 to 52 (test-time scaling) (ablation)OmniAgent · VideoMME-Longaccuracy: 53.4 -> 59.6 (+6.2%)Improved
use active on-demand perception instead of dense passive sampling (vs Qwen2.5-VL-72B)OmniAgent-7B · LVBenchaccuracy: 47.3 -> 50.5 with 203 vs 768 framesImproved
longer videos (about 30 to 130 min) with active perception (ablation)OmniAgent · LVBench by durationaccuracy: 53.7 (20-40 min) -> 50.8 (120-140 min)No clear effect

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