# One Agent to Serve All: a Lite-Adaptive Stylized AI Assistant for Millions of Multi-Style Official Accounts

arXiv: https://arxiv.org/abs/2509.17788 · HTML: https://mlsanity.com/paper/2509.17788/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 10 of its 20 references that are in the map.
- First posted on arXiv 22 Sep 2025.

## Builds on (most cited first, 10 of 10)

- [ChatQA: Surpassing GPT-4 on Conversational QA and RAG](https://mlsanity.com/paper/2401.10225/) (2024)
- [A Dual Reinforcement Learning Framework for Unsupervised Text Style Transfer](https://mlsanity.com/paper/1905.10060/) (2019)
- [Structuring Latent Spaces for Stylized Response Generation](https://mlsanity.com/paper/1909.05361/) (2019)
- [Stylized Dialogue Response Generation Using Stylized Unpaired Texts](https://mlsanity.com/paper/2009.12719/) (2020)
- [Are Large Language Models Actually Good at Text Style Transfer?](https://mlsanity.com/paper/2406.05885/) (2024)
- [Non-Parallel Text Style Transfer with Self-Parallel Supervision](https://mlsanity.com/paper/2204.08123/) (2022)
- [Stylized Data-to-Text Generation: A Case Study in the E-Commerce Domain](https://mlsanity.com/paper/2305.03256/) (2023)
- [Customizing Large Language Model Generation Style using Parameter-Efficient Finetuning](https://mlsanity.com/paper/2409.04574/) (2024)
- [StyleChat: Learning Recitation-Augmented Memory in LLMs for Stylized Dialogue Generation](https://mlsanity.com/paper/2403.11439/) (2024)
- [Stylized Knowledge-Grounded Dialogue Generation via Disentangled Template Rewriting](https://mlsanity.com/paper/2204.05610/) (2022)

## Built on by (most cited first, 1 of 1)

- [Semantic Flow Regularization: Teaching LLMs to Generate Diverse Yet Coherent Responses](https://mlsanity.com/paper/2605.27971/) (2026)

## Abstract (arXiv)

Conversational agents deployed in industrial-scale official account platforms must generate responses that are both contextually grounded and stylistically aligned-requirements that existing methods struggle to meet. Chain-of-thought (CoT) prompting induces significant latency due to multi-turn reasoning; per-account fine-tuning is computationally prohibitive; and long prompt-based methods degrade the model's ability to grasp injected context and style. In this paper, we propose WeStar, a lite-adaptive framework for stylized contextual question answering that scales to millions of official accounts. WeStar combines context-grounded generation via RAG with style-aware generation using Parametric RAG (PRAG), where LoRA modules are dynamically activated per style cluster. Our contributions are fourfold: (1) We introduce WeStar, a unified framework capable of serving large volumes of official accounts with minimal overhead. (2) We propose a multi-dimensional, cluster-based parameter sharing scheme that enables compact style representation while preserving stylistic diversity. (3) We develop a style-enhanced Direct Preference Optimization (SeDPO) method to optimize each style cluster's parameters for improved generation quality. (4) Experiments on a large-scale industrial dataset validate the effectiveness and efficiency of WeStar, underscoring its pracitical value in real-world deployment.