HUOZIIME: An On-Device LLM-enhanced Input Method for Deep Personalization
This work addresses the challenge of personalized text input on mobile devices, offering a privacy-preserving solution that adapts to user-specific history in real-time.
HUOZIIME introduces an on-device LLM-powered input method that achieves deep personalization through hierarchical memory and post-training on synthetic data, enabling efficient and privacy-preserving text prediction on mobile devices.
Mobile input method editors (IMEs) are the primary interface for text input, yet they remain constrained to manual typing and struggle to produce personalized text. While lightweight large language models (LLMs) make on-device auxiliary generation feasible, enabling deeply personalized, privacy-preserving, and real-time generative IMEs poses fundamental challenges.To this end, we present HUOZIIME, a personalized on-device IME powered by LLM. We endow HUOZIIME with initial human-like prediction ability by post-training a base LLM on synthesized personalization data. Notably, a hierarchical memory mechanism is designed to continually capture and leverage user-specific input history. Furthermore, we perform systemic optimizations tailored to on-device LLMbased IME deployment, ensuring efficient and responsive operation under mobile constraints.Experiments demonstrate efficient on-device execution and high-fidelity memory-driven personalization. Code and package are available at https://github.com/Shan-HIT/HuoziIME.