跳到正文
HuggingFace Daily Papers(社區熱門論文)·· 2026-10-07AI 評分40

Less from More: Reinforcing Sparse Video Reasoning from Dense References

Less from More: Reinforcing Sparse Video Reasoning from Dense References

AI 導讀

SAVER 提出由密集到稀疏的後訓練框架,以密集影片視角作為訓練參照,提升有限影格下的影片推理可靠性。模型只用 1,250 個隨機抽樣的時間定位例子訓練,無須影片問答標註。SAVER 在 3 個時間定位基準及 6 個影片問答基準上均有提升,使用大幅較少影格時仍可匹敵或超越密集影格 Qwen3.5 基線。

正文

Published on Oct 7

Authors:

,

Abstract

Video-language models commonly assume that more temporal observations lead to more reliable reasoning. We question this assumption and argue that the key challenge is not merely processing more video frames efficiently, but learning to reason reliably under limited temporal evidence. We propose SAVER, a dense-to-sparse post-training framework that uses dense video views as training-time references for sparse-frame inference. During reinforcement post-training, paired dense and sparse views are optimized with grounding rewards and a reliability-gated reference reward, encouraging sparse view predictions to preserve task-relevant temporal evidence. Notably, SAVER is trained only on 1,250 randomly sampled temporal grounding examples, without using any video question answering annotations. Across three temporal grounding benchmarks and six video question-answering benchmarks, SAVER consistently improves performance across frame budgets. In particular, SAVER can match or surpass dense-frame Qwen3.5 baselines while using substantially fewer frames. These results show that temporal grounding can serve as an effective evidence-localization proxy for learning sparse video reasoning that transfers to broader video understanding tasks.

View arXiv page View PDF Project page GitHub Add to collection

Get this paper in your agent:

hf papers read 2610.10893

Don't have the latest CLI?

curl -LsSf https://hf.co/cli/install.sh | bash

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2610.10893 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2610.10893 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2610.10893 in a Space README.md to link it from this page.

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.

來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co