cs.CVNov 10, 2026

EVAThink: Evidence Sufficiency-Guided Training for Adaptive Thinking in Video Understanding

Authors: Changti Wu, Ruizhe Fan, Bin Wang, Lei Zhang, Yuxiang Zhang

Organizations: East China Normal University · Zhongguancun Academy · Huawei Technologies Co., Ltd

Abstract

Under a fixed frame budget, video understanding with vision-language models (VLMs) depends on whether the sampled frames provide sufficient evidence for the question. Sufficient frame evidence supports direct answering, whereas less informative observations may benefit from reasoning that connects the available cues. We introduce EVAThink, which enhances video understanding through evidence sufficiency-guided training for adaptive thinking. First, Recoverable Evidence Gap Mining constructs evidence-dense and evidence-sparse video views that share the same question and answer under an identical frame budget. It retains pairs where sparse views yield lower direct-answer accuracy than dense views but benefit from explicit reasoning, providing supervision for response-mode selection. Second, Evidence-Guided Two-Stage Reinforcement Learning uses stage-specific rewards to guide training with evidence-based mode preferences and then refine response efficiency, while first-token advantage regularization mitigates mode collapse. Trained on only 6.7K curated samples, EVAThink consistently improves upon the base model across video understanding benchmarks under both 32- and 64-frame budgets, while outperforming strong baselines on average. The code, model, and datasets are available at EVAThink.

Figures & tables

Appendix figures & tables12 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. Video-Zero: Self-Evolution Video Understanding

    May 14, 2026Ruixu Zhang, Deyi Ji, Lanyun Zhu +4Temporal Video GroundingLong-Video Understanding

  2. Watch Before You Answer: Learning from Visually Grounded Post-Training

    Apr 6, 2026Yuxuan Zhang, EunJeong Hwang, Huaisong Zhang +8VLM Post-TrainingVision-Language Models

  3. When and Where to Look: Adaptive Visual Evidence Scheduling for Efficient Long Video Understanding

    Aug 4, 2026Ke Li, Jiayu Chen, Maoliang Li +5Efficient VLM InferenceLarge Vision-Language Models