arXiv cs.AI· Grayson Matthew, Lokesh Boominathan, Yizhou Chen, Matthew McGinley, Xaq Pitkow·· 9 天前AI 评分22
何时需要注意力：资源高效智能体的最优注意力分配策略
Attention when you need
AI 导读
一项基于信号检测任务的规范模型研究揭示了资源高效智能体的最优注意力分配策略：当注意力成本较高时，最优注意力随信号证据积累逐步攀升；当注意力成本较低时，最优注意力呈节律性波动，且波动频率随奖励幅度、信号短暂性和信号频率等任务参数增加而提高。研究认为，节律性注意力自然涌现于信念更新由稳定时间先验主导、而非新感觉观测似然主导的智能体。
正文 · 原文
Abstract:Paying attention improves performance, but attention is metabolically costly, so how should a resource-efficient agent allocate it? We study optimal allocation strategies using a normative model of a signal detection task in which attention comes at a cost. The model reveals that optimal attention is temporally structured in one of two patterns depending on task conditions: when attention costs penalize intense focus, optimal attention ramps up as evidence for the signal accumulates, but with less prohibitive attention costs, optimal attention fluctuates rhythmically. In cases with rhythmic attention, allocation frequency increases with task parameters such as reward magnitude for successful detection, signal brevity, and signal frequency. We argue that rhythmic attention emerges naturally in an agent whose belief updates are dominated by a stable temporal prior and not by the likelihood of new sensory observations. Our results characterize the conditions under which rhythmic vs. ramping attention is optimal and offer a normative account of attentional fluctuations observed in sustained attention tasks.
| Subjects: | Neurons and Cognition (q-bio.NC); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2501.07440 [q-bio.NC] |
| (or arXiv:2501.07440v3 [q-bio.NC] for this version) | |
| https://doi.org/10.48550/arXiv.2501.07440 arXiv-issued DOI via DataCite |
Submission history
From: Lokesh Boominathan [view email]
[v1]
Mon, 13 Jan 2025 16:08:47 UTC (2,496 KB)
[v2]
Wed, 29 Jan 2025 16:04:17 UTC (2,496 KB)
[v3]
Thu, 8 Oct 2026 06:19:20 UTC (1,452 KB)
来源：arXiv cs.AI · arxiv.org