AuthorMix: Modular Authorship Style Transfer via Layer-wise Adapter Mixing

• URL: http://arxiv.org/abs/2603.23069v1
• Date: Tue, 24 Mar 2026 11:07:00 GMT
• Title: AuthorMix: Modular Authorship Style Transfer via Layer-wise Adapter Mixing
• Abstract summary: AuthorMix is a lightweight, modular, and interpretable style transfer framework.<n>We train individual, style-specific LoRA adapters on a small set of high-resource authors.
• Score: 50.94046654012963
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The task of authorship style transfer involves rewriting text in the style of a target author while preserving the meaning of the original text. Existing style transfer methods train a single model on large corpora to model all target styles at once: this high-cost approach offers limited flexibility for target-specific adaptation, and often sacrifices meaning preservation for style transfer. In this paper, we propose AuthorMix: a lightweight, modular, and interpretable style transfer framework. We train individual, style-specific LoRA adapters on a small set of high-resource authors, allowing the rapid training of specialized adaptation models for each new target via learned, layer-wise adapter mixing, using only a handful of target style training examples. AuthorMix outperforms existing, SoTA style-transfer baselines -- as well as GPT-5.1 -- for low-resource targets, achieving the highest overall score and substantially improving meaning preservation.

Related papers

• StyleAdaptedLM: Enhancing Instruction Following Models with Efficient Stylistic Transfer [4.077787659104315]
  StyleAdaptedLM is a framework that efficiently transfers stylistic traits to instruction-following models.<n>LoRA adapters are first trained on a base model with diverse unstructured stylistic corpora, then merged with a separate instruction-following model.
  arXiv  Detail & Related papers  (2025-07-24T10:57:32Z)
• Authorship Style Transfer with Policy Optimization [26.34892894935038]
  Authorship style transfer aims to rewrite a given text into a specified target while preserving the original meaning in the source. Existing approaches rely on the availability of a large number of target style exemplars for model training.
  arXiv  Detail & Related papers  (2024-03-12T19:34:54Z)
• HiCAST: Highly Customized Arbitrary Style Transfer with Adapter Enhanced Diffusion Models [84.12784265734238]
  The goal of Arbitrary Style Transfer (AST) is injecting the artistic features of a style reference into a given image/video. We propose HiCAST, which is capable of explicitly customizing the stylization results according to various source of semantic clues. A novel learning objective is leveraged for video diffusion model training, which significantly improve cross-frame temporal consistency.
  arXiv  Detail & Related papers  (2024-01-11T12:26:23Z)
• STEER: Unified Style Transfer with Expert Reinforcement [71.3995732115262]
  STEER: Unified Style Transfer with Expert Reinforcement, is a unified frame-work developed to overcome the challenge of limited parallel data for style transfer. We show STEER is robust, maintaining its style transfer capabilities on out-of-domain data, and surpassing nearly all baselines across various styles.
  arXiv  Detail & Related papers  (2023-11-13T09:02:30Z)
• ParaGuide: Guided Diffusion Paraphrasers for Plug-and-Play Textual Style Transfer [57.6482608202409]
  Textual style transfer is the task of transforming stylistic properties of text while preserving meaning. We introduce a novel diffusion-based framework for general-purpose style transfer that can be flexibly adapted to arbitrary target styles. We validate the method on the Enron Email Corpus, with both human and automatic evaluations, and find that it outperforms strong baselines on formality, sentiment, and even authorship style transfer.
  arXiv  Detail & Related papers  (2023-08-29T17:36:02Z)
• Master: Meta Style Transformer for Controllable Zero-Shot and Few-Shot Artistic Style Transfer [83.1333306079676]
  In this paper, we devise a novel Transformer model termed as emphMaster specifically for style transfer. In the proposed model, different Transformer layers share a common group of parameters, which (1) reduces the total number of parameters, (2) leads to more robust training convergence, and (3) is readily to control the degree of stylization. Experiments demonstrate the superiority of Master under both zero-shot and few-shot style transfer settings.
  arXiv  Detail & Related papers  (2023-04-24T04:46:39Z)
• Low-Resource Authorship Style Transfer: Can Non-Famous Authors Be Imitated? [41.365967145680116]
  Authorship style transfer involves altering text to match the style of a target author whilst preserving the original meaning. We introduce the low-resource authorship style transfer task, where only a limited amount of text in the target author's style may exist. In our experiments, we specifically choose source and target authors from Reddit and style transfer their Reddit posts, limiting ourselves to just 16 posts (on average 500 words) of the target author's style.
  arXiv  Detail & Related papers  (2022-12-18T01:57:30Z)
• Motion Style Transfer: Modular Low-Rank Adaptation for Deep Motion Forecasting [79.56014465244644]
  We propose a transfer learning approach for efficiently adapting deep motion forecasting models to new domains. Unlike the conventional fine-tuning approach that updates the whole encoder, our main idea is to reduce the amount of tunable parameters. We show that our proposed adapter design, coined MoSA, outperforms prior methods on several forecasting benchmarks.
  arXiv  Detail & Related papers  (2022-11-06T16:14:17Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.