ResearchResearch paperLarge Language Models · Training & Scaling · Efficiency & Inference1 source · Oct 20, 2026

PACE: Plan, Attend, Check, and Emit for Training-Free Response Timing in Streaming Video Understanding

Streaming video understanding requires a model to decide not only what to answer but also when to answer.

Existing approaches often learn response timing through additional training, while training-free alternatives typically rely on scalar gating signals that do not explicitly represent when an answer becomes final.
We organize chunk-wise streaming errors into three patterns: sampling variation can make an answer unstable, a stable answer may lack support in the observed video, or an answer may be supported by the current evidence while the event remains unfinished.
We propose PACE (Plan, Attend, Check, and Emit), a training-free framework designed around these three failure modes.
Experiments on ProactiveVideoQA and OVO-Bench Forward show that PACE improves over the official protocols by up to 5.9 average PAUC points and 6.2 average accuracy points across proprietary and open-source backbones, yielding state-of-the-art performance.

[1]PACE: Plan, Attend, Check, and Emit for Training-Free Response Timing in Streaming Video Understanding
arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 20, 02:58 PM
Streaming video understanding requires a model to decide not only what to answer but also when to answer.
Existing approaches often learn response timing through additional training, while training-free alternatives typically rely on scalar gating signals that do not explicitly represent when an answer becomes final.

Extractive summary: sentences quoted from the sources.

Before this