風格重組：透過風格元素的提煉和擾動實現可解釋的作者身份混淆

摘要

作者身份混淆是一項重要但具有挑戰性的任務，指的是將文本重新編寫，以故意掩蓋作者的身份。目前使用大型語言模型（LLMs）的方法缺乏可解釋性和可控性，通常忽略特定作者的風格特徵，導致整體性能較差。為了應對這一問題，我們開發了StyleRemix，這是一種適應性和可解釋性的混淆方法，它會干擾原始輸入文本的特定、細粒度的風格元素。StyleRemix 使用預先訓練的低秩適應（LoRA）模塊來重新編寫輸入文本，沿著各種風格軸（例如正式性和長度）進行調整，同時保持低計算成本。在自動和人工評估中，StyleRemix 在各種領域中均優於最先進的基準和更大的LLMs。此外，我們還發布了AuthorMix，這是一個包含30,000篇高質量長文本的大型數據集，來自14位作者和4個領域，以及DiSC，這是一個包含1,500篇文本的平行語料庫，涵蓋了16個獨特方向上的七個風格軸。

English

Authorship obfuscation, rewriting a text to intentionally obscure the identity of the author, is an important but challenging task. Current methods using large language models (LLMs) lack interpretability and controllability, often ignoring author-specific stylistic features, resulting in less robust performance overall. To address this, we develop StyleRemix, an adaptive and interpretable obfuscation method that perturbs specific, fine-grained style elements of the original input text. StyleRemix uses pre-trained Low Rank Adaptation (LoRA) modules to rewrite an input specifically along various stylistic axes (e.g., formality and length) while maintaining low computational cost. StyleRemix outperforms state-of-the-art baselines and much larger LLMs in a variety of domains as assessed by both automatic and human evaluation. Additionally, we release AuthorMix, a large set of 30K high-quality, long-form texts from a diverse set of 14 authors and 4 domains, and DiSC, a parallel corpus of 1,500 texts spanning seven style axes in 16 unique directions

風格重組：透過風格元素的提煉和擾動實現可解釋的作者身份混淆

StyleRemix: Interpretable Authorship Obfuscation via Distillation and Perturbation of Style Elements

摘要

Support