Fints: Efficient Inference-Time Personalization for LLMs with Fine-Grained Instance-Tailored Steering

TL;DR

Fints introduces a dynamic, fine-grained personalization method for LLMs that injects instance-specific signals during inference, improving adaptability and data efficiency in rapidly changing environments.

Contribution

It proposes a novel, plug-in steering framework that dynamically generates sample-level interference vectors for personalized LLM adaptation, compatible with existing methods.

Findings

Significantly improves personalization in fast-changing scenarios.

Enhances performance with high data sparsity.

Maintains robustness across different interaction modes.

Abstract

The rapid evolution of large language models (LLMs) has intensified the demand for effective personalization techniques that can adapt model behavior to individual user preferences. Despite the non-parametric methods utilizing the in-context learning ability of LLMs, recent parametric adaptation methods, including personalized parameter-efficient fine-tuning and reward modeling emerge. However, these methods face limitations in handling dynamic user patterns and high data sparsity scenarios, due to low adaptability and data efficiency. To address these challenges, we propose a fine-grained and instance-tailored steering framework that dynamically generates sample-level interference vectors from user data and injects them into the model's forward pass for personalized adaptation. Our approach introduces two key technical innovations: a fine-grained steering component that captures nuanced signals by hooking activations from attention and MLP layers, and an input-aware aggregation module that synthesizes these signals into contextually relevant enhancements. The method demonstrates high flexibility and data efficiency, excelling in fast-changing distribution and high data sparsity scenarios. In addition, the proposed method is orthogonal to existing methods and operates as a plug-in component compatible with different personalization techniques. Extensive experiments across diverse scenarios--including short-to-long text generation, and web function calling--validate the effectiveness and compatibility of our approach. Results show that our method significantly enhances personalization performance in fast-shifting environments while maintaining robustness across varying interaction modes and context lengths. Implementation is available at https://github.com/KounianhuaDu/Fints.