RPM: Reasoning-Level Personalization for Black-Box Large Language Models

TL;DR

This paper introduces RPM, a novel framework for black-box large language models that personalizes responses by modeling user-specific reasoning structures from behavioral data, improving both accuracy and interpretability.

Contribution

RPM is the first systematic approach to discover user reasoning paths and guide inference, moving beyond response-level personalization to reasoning-level personalization.

Findings

RPM outperforms existing methods across four tasks.

It enhances personalization performance.

It improves interpretability of model responses.

Abstract

While black-box large language models are widely deployed, they produce generic outputs that overlook individual user preferences. Current personalization methods are fundamentally limited to response-level personalization; they only match final outputs, failing to model the underlying reasoning that connects user behavior to responses. To address this, this work introduces reasoning-level personalization as a new paradigm and proposes RPM, the first systematic framework that automatically discovers user-specific reasoning structures from raw behavioral data to guide the model's personalized inference. RPM constructs a structured model of user behavior-built from response-influential features and statistical factors-to create personalized reasoning paths and retrieve beneficial examples for guiding inference through a feature-based retrieval mechanism. Extensive experiments across four diverse tasks demonstrate that RPM consistently outperforms existing response-level methods while simultaneously enhancing both personalization performance and interpretability, providing a promising direction for black-box LLM personalization.