---
title: "Persona Learning from Feedback for Personalized AI Assistants"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2610.38561v1"
permalink: "https://feed.astralor.com/items/itm_276e13b7897e1018ac91658c"
published: "2026-10-28T09:55:26.000Z"
score: 55
channel: "tech"
tags: ["多模态","Persona Learning","Feedback","Personalized","Assistants Personalizing","Existing","Personalizing","However"]
---

# Persona Learning from Feedback for Personalized AI Assistants

## At a glance

Personalizing AI assistants requires adapting their behavior to users' preferences and task contexts. Existing work elicits assistant personas through persona prompting or finetuning on persona-conditioned generations. …

## Key points

- Personalizing AI assistants requires adapting their behavior to users' preferences and task contexts.
- Existing work elicits assistant personas through persona prompting or finetuning on persona-conditioned generations.
- However, persona descriptions alone do not identify what to change or preserve in a particular response, or which behaviors suit the current request.

## Full text

Personalizing AI assistants requires adapting their behavior to users' preferences and task contexts. Existing work elicits assistant personas through persona prompting or finetuning on persona-conditioned generations. However, persona descriptions alone do not identify what to change or preserve in a particular response, or which behaviors suit the current request. We introduce $\textit{Persona Learning from Feedback (PLF)}$, a framework that learns reusable, preference-grounded assistant personas as independent LoRA adapters on a shared frozen backbone. During training, a frozen self-teacher conditions on persona specifications and response-specific feedback to provide token-level supervision on each adapter's own generated prefixes. A feedback gate focuses corrective updates on responses judged to need improvement. At deployment, a supervised router uses the request or dialogue to select a single adapter or a fixed composition, without user-specific finetuning. We evaluate PLF on single-turn personalization with model-generated critiques and multi-turn tutoring with simulated students' reactions. Across both settings, PLF improves model-judged preference fit over fixed-persona policies and strong baselines including persona prompting and persona distillation, demonstrating how feedback-trained behaviors can be reused for adaptive assistant personalization without degrading general capabilities.

---

Verify against the original before citing; curated by Astral Feed: https://feed.astralor.com/items/itm_276e13b7897e1018ac91658c