发表机构
University of Stavanger（斯塔万格大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
InMyStyle是注重隐私的单用户系统，用多本地辅助LLM构建配对训练示例，微调0.5B-7B参数模型的LoRA适配器，可将AI编辑文本重写为用户风格，小模型即可完成重写任务，其输出感知AI性低于辅助AI生成内容
AI 中文摘要
InMyStyle是一个注重隐私的单用户系统，可适配小型语言模型，在推理时无需指令提示即可将AI编辑的文本重写为符合单个用户写作风格的内容。给定用户的文档，该系统利用多个本地辅助LLM构建配对训练示例，并对参数规模在0.5B至7B的基础模型微调LoRA适配器。长度感知生成预算与自动分块功能支持不同长度的输入。在来自科学论文语料库的219个评估对上，无论是贪心解码还是采样解码，所有模型规模的自动复合分数都稳定在0.69（0-1分制）。该观察到的稳定状态表明，小型模型足以完成所测重写任务，模型规模决定权衡因素而非稳定的质量排名。作为二次评估，来自5名LLM评判者的400项评分显示，InMyStyle输出的平均感知AI性分数比其辅助AI生成的输入低20%以上，且InMyStyle内部的平均感知AI性分数随模型规模增大而降低。
英文摘要
InMyStyle is a privacy-first, single-user system that adapts small language models to rewrite AI-edited text towards an individual user's writing style without an instruction prompt at inference. Given a user's documents, it uses multiple local helper LLMs to construct paired training examples and fine-tunes LoRA adapters on Qwen2.5 models ranging from 0.5B to 7B parameters. Length-aware generation budgets and automatic chunking support inputs of different lengths. We report a single-user case study: 219 evaluation pairs derived from 73 paragraphs of one author's scientific writing, with all adapters trained using the same rank-8, three-epoch recipe. The automatic composite score (0-1 scale) plateaus across model sizes under both greedy and sampled decoding ($Q=0.689$-$0.695$, with overlapping confidence intervals). In this setting, small models are sufficient for the measured rewriting task, and model size mainly determines efficiency trade-offs rather than a stable quality ranking. The gains favor content-preserving naturalization more than recovery of personal style, with authorship probabilities staying near the classifier's decision boundary (0.51--0.53) and stylometric improvement being near zero. As a secondary evaluation, 400 ratings from five LLM judges give InMyStyle outputs a mean perceived AI-ness score over 20% lower than their helper-generated inputs, with scores decreasing with model size in this sample. The study does not establish generalization across users.
