An Exam for Active Observers
- 类型：arxiv
- 标识：2607.16165
- 链接：https://arxiv.org/abs/2607.16165
- 主分类：multimodal
- 形态：benchmark
- 被引：0
- 被引来源：Semantic Scholar + OpenAlex
- S2被引：0
- OpenAlex被引：0
- 影响力被引：0
- TLDR：Human vision is a closed loop: gaze is continuously redirected by intermediate hypotheses rather than a single snapshot. Decades of psychophysics and cognitive science have argued that this active observation is essential for a wide range of tasks. Whether today's multimodal large language models (MLLMs) exercise active observation is an empirical question that current vision-language benchmarks do not answer. We introduce ActiveVision, a benchmark that makes active observation measurable for MLLMs, comprising 17 tasks across 3 categories. Tasks are designed to force repeated visual perception
- OpenAlex ID：W7169754001
- OpenAlex DOI：10.48550/arxiv.2607.16165
- DOI：10.48550/arxiv.2607.16165
- DOI来源：OpenAlex
- 开放获取：green
- 开放获取链接：https://doi.org/10.48550/arxiv.2607.16165
- OpenAlex更新：2026-08-24
- 副分类：evaluation
- 待LLM分类：否
- 标题中文：面向主动观察者的评测
- TLDR中文：人类视觉是一个闭环：注视点不断被中间假设而非单一快照持续重定向。数十年的心理物理学与认知科学研究表明，主动观察对多种任务至关重要。当代多模态大语言模型 (MLLM) 是否进行主动观察，是一个现有视觉语言基准无法回答的经验问题。我们提出 ActiveVision，一个使 MLLM 主动观察可度量的基准，包含 3 个类别共 17 个任务，任务设计强制进行重复视觉感知……
- 来源文件：
- /inbox/tom/_candidates/2026-07-23-agent-rag-longcontext-candidates.json
- /inbox/tom/_candidates/2026-07-24-agent-rag-longcontext-candidates.json
- [S2 enrich]
- [OpenAlex backfill]