Hypotheses-Guided Self Distillation for Continual Personalization

2026-08-31

AI Summary

为解决持续个性化问题，提出HypReflect框架，通过从多样用户信号中推断偏好假设并进行自我蒸馏来实现有效长期交互。

Abstract

As people increasingly interact with LLM assistants in daily life, continually adapting to individual preferences has become essential for effective long-term interactions. However, user preferences are rarely stated in full, and instead emerge through heterogeneous, latent, and noisy signals, with existing methods relying on raw interaction histories or costly reward-based optimization to manage personalization. We introduce HypReflect, a reliable, scalable framework for continual personalization that infers explicit, uncertainty-aware preference hypotheses from diverse user signals, reflectively refines them as new evidence accumulates, and incorporates the resulting user model through hypotheses-guided self-distillation. Experiments across three personalization settings: online personalization, multi-session interactions, and implicit behavioral signals, show that HypReflect outperforms a range of baselines, including raw-history and incremental-update methods. We further demonstrate strong generalization to unseen users and cross-domain settings, along with stability across context budgets, reusable hypotheses, and more focused personalization. These results suggest a step towards reliable and scalable continual personalization through explicit, revisable user preference hypotheses.

Problem

• Continual Personalization
• User Preferences
• Heterogeneous Signals
• Latent Signals
• Noisy Signals

Innovation

• Continual Personalization
• Hypotheses-Guided Self Distillation
• Uncertainty-Aware Preference Hypotheses
• Scalable Framework