ProactiveMobile: A Comprehensive Benchmark for Boosting Proactive Intelligence on Mobile Devices
By: Dezhi Kong, Zhengzhao Feng, Qiliang Liang, Hao Wang, Haofei Sun, Changpeng Yang, Yang Li, Peng Zhou, Shuai Nie, Hongzhen Wang, Linfeng Zhou
Published: 2026-02-26
View on arXiv →#cs.AI
Abstract
ProactiveMobile introduces a new, comprehensive benchmark for evaluating and advancing proactive intelligence on mobile devices. It focuses on scenarios where AI anticipates user needs and provides timely assistance without explicit prompts, covering tasks from personalized notifications to intelligent resource management, thereby pushing the boundaries of on-device AI efficiency and user experience. This has direct real-world application in developing next-generation smartphone features, wearable technology, and smart assistants.