HUOZIIME: An On-Device LLM-enhanced Input Method for Deep Personalization

TL;DR

HUOZIIME is an on-device input method that leverages a lightweight LLM with a hierarchical memory mechanism to provide personalized, privacy-preserving, and real-time text input on mobile devices.

Contribution

The paper introduces HUOZIIME, a novel personalized on-device IME that combines a post-trained LLM with a hierarchical memory system and system optimizations for mobile deployment.

Findings

Efficient on-device execution demonstrated.

High-fidelity user-specific personalization achieved.

Systematic optimizations enable real-time responsiveness.

Abstract

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.