Instant Personalized Large Language Model Adaptation via Hypernetwork
This addresses the scalability and real-time update challenges in personalized LLMs for large-scale applications, representing a novel method rather than an incremental improvement.
The paper tackles the computational inefficiency of training separate adapters for each user in personalized LLMs by introducing Profile-to-PEFT, a hypernetwork-based framework that maps user profiles to adapter parameters, eliminating per-user training and enabling instant adaptation. Experimental results show it outperforms existing methods like prompt-based personalization and OPPU while using fewer resources, with strong generalization to unseen users and robustness across different conditions.
Personalized large language models (LLMs) tailor content to individual preferences using user profiles or histories. However, existing parameter-efficient fine-tuning (PEFT) methods, such as the ``One-PEFT-Per-User'' (OPPU) paradigm, require training a separate adapter for each user, making them computationally expensive and impractical for real-time updates. We introduce Profile-to-PEFT, a scalable framework that employs a hypernetwork, trained end-to-end, to map a user's encoded profile directly to a full set of adapter parameters (e.g., LoRA), eliminating per-user training at deployment. This design enables instant adaptation, generalization to unseen users, and privacy-preserving local deployment. Experimental results demonstrate that our method outperforms both prompt-based personalization and OPPU while using substantially fewer computational resources at deployment. The framework exhibits strong generalization to out-of-distribution users and maintains robustness across varying user activity levels and different embedding backbones. The proposed Profile-to-PEFT framework enables efficient, scalable, and adaptive LLM personalization suitable for large-scale applications.