This research explores strategies for steering the output of large language models (LLMs) towards specific styles, such as sentiment, emotion, or writing style, by adding style vectors to the activations of hidden layers during text generation. We show that style vectors can be simply computed from recorded layer activations for input texts in a specific style in contrast to more complex training-based approaches. Through a series of experiments, we demonstrate the effectiveness of activation engineering using such style vectors to influence the style of generated text in a nuanced and parameterisable way, distinguishing it from prompt engineering. The presented research constitutes a significant step towards developing more adaptive and effective AI-empowered interactive systems.
Style-Specific Neurons for Steering LLMs in Text Style Transfer
School of Computation, Information and Technology, Technical University of Munich（慕尼黑工业大学计算、信息与技术学院）；Munich Center for Machine Learning（慕尼黑机器学习中心）；Center for Information and Language Processing, LMU Munich（慕尼黑大学信息与语言处理中心）
Steering at the Source: Style Modulation Heads for Robust Persona Control
The University of Tokyo, Tokyo, Japan（东京大学）；National Institute of Informatics Research（信息处理研究所）；Development Center for Large Language Models, Japan（大型语言模型发展中心）
Overview of PAN 2026: Voight-Kampff Generative AI Detection, Text Watermarking, Multi-Author Writing Style Analysis, Generative Plagiarism Detection, and Reasoning Trajectory Detection
Friedrich Schiller University Jena（耶纳弗里德里希·施勒格尔大学）；University of Applied Sciences BFI（应用科学大学BFI）；Mohamed bin Zayed University of Artificial Intelligence（马尔代夫 bin Zayed 人工智能大学）；University of Kassel（卡塞尔大学）；INSAIT, Sofia University ``St. Kliment Ohridski''（INSAIT，索菲亚大学『圣克莱门特·奥赫里迪斯』）；University of Innsbruck（因斯布鲁克大学）
PAN 2026旨在通过客观评估推进计算风格学和文本取证，涵盖生成式AI检测、文本水印、多作者写作分析、抄袭检测及推理轨迹检测等五个新旧任务。
VIS Lab, University of Amsterdam（阿姆斯特丹大学VIS实验室）；FunAI Lab, University of Technology Nuremberg（纽伦堡技术大学FunAI实验室）；CSAIL, Massachusetts Institute of Technology（麻省理工学院计算机科学与人工智能实验室）
The Invisible Editorial Layer: Formalizing Undisclosed Inference-Time Steering, Probability Placement, and the Attribution Problem in Deployed Language Models
Personalized Text Generation with Contrastive Activation Steering
Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）；School of Artificial Intelligence, University of Chinese Academy of Sciences（中国科学院大学人工智能学院）；Northeastern University（东北大学）；University of Science and Technology of China（中国科学技术大学）；National University of Singapore（新加坡国立大学）
NTT Media Intelligence Laboratory, NTT Corporation（日本电信电话株式会社媒体智能实验室）；The University of Tokyo（东京大学）
本研究探讨了生成性阅读理解（RC），它包括基于文本证据的回答问题和自然语言生成（NLG）。我们提出了一种多风格的抽象摘要模型的问答，称为面具。所提出的模型有两个关键特征。首先，与大多数关于RC的研究不同，这些研究专注于从提供的段落中提取答案范围，我们的模型专注于从问题和多个段落中生成摘要。这涵盖了现实世界应用程序所需的各种回答风格。其次，由于难以获得一个通用模型，因此以前的研究为每种回答风格建立了一个特定的模型，而我们的方法在一个模型中学习多风格的答案，以提高所有类型的NLG能力。这也使我们的模型能够以目标风格给出答案。实验结果表明，该模型在MS MARCO 2.1的Q&A任务和Q&A + NLG任务以及NarrativeQA的摘要任务上都达到了最好的性能。我们观察到，转移的风格独立的NLG能力的目标风格是其成功的关键。
Unsupervised Text Style Transfer with Deep Generative Models
Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）；School of Artificial Intelligence, University of Chinese Academy of Sciences（中国科学院大学人工智能学院）
Multi-span Style Extraction for Generative Reading Comprehension
SJTU-ParisTech Elite Institute of Technology, Shanghai Jiao Tong University（上海交通大学密西根学院（SJTU-ParisTech Elite Institute of Technology））；Department of Computer Science and Engineering, Shanghai Jiao Tong University（上海交通大学计算机科学与工程系）；Key Laboratory of Shanghai Education Commission for Intelligent Interaction and Cognitive Engineering, Shanghai Jiao Tong University（上海交通大学上海教委智能交互与认知工程重点实验室）；MoE Key Lab of Artificial Intelligence, AI Institute, Shanghai Jiao Tong University（上海交通大学人工智能研究院教育部人工智能重点实验室）
