LoRA-generating hypernetworks for efficient on-device LLM generative personalization

Sean Augenstein, Li Ding, Jihwan Lee, Keith Rush, Andrey Zhmoginov|September 21, 2026arXiv

Key Takeaway

Hypernetworks can efficiently generate personalized LoRA adapters on mobile devices by mapping user context to weight modifications, avoiding the latency costs of context extension while remaining computationally feasible for resource-constrained devices.

Summary

This paper presents a method for personalizing language models on mobile devices by training a hypernetwork that generates customized LoRA (low-rank adaptation) weights based on a user's context.

efficiency applications

Key Terms

low-rank-adaptation hypernetwork in-context-learning parameter-efficient-fine-tuning personalization