English(EN) GridProbe: Posterior-Probing for Adaptive Test-Time Compute in Long-Video VLMs

GridProbe 降低长视频 VLM 的计算成本

作者 PulseAugur 编辑部 · [1 个来源] · 2026-05-11 15:57

研究人员开发了 GridProbe，一种提高长视频视觉语言模型 (VLM) 效率的新方法。该技术在推理过程中自适应地选择相关帧，从而降低了处理数千帧的计算成本。GridProbe 通过在答案空间中探测帧的重要性来实现这一点，从而能够在不牺牲准确性的情况下根据问题难度动态调整处理的帧数。 AI

影响降低了使用 AI 处理长视频内容的计算需求，可能有助于更广泛地采用 VLM 应用。

排序理由发布了一篇详细介绍提高 AI 模型效率新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 0 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 发布了一篇详细介绍提高 AI 模型效率新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, infra; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: 183 days old; Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.

完整方法见我们的编辑标准。

arXiv cs.CV TIER_1 English(EN) · Naeemullah Khan · 2026-05-11 15:57

GridProbe：长视频VLMs中用于自适应测试时计算的后验探测

Long-video understanding in VLMs is bottlenecked by a single monolithic forward pass over thousands of frames at quadratic attention cost. A common mitigation is to first select a small subset of informative frames before the forward pass; common for training-free selectors via a…