The paper investigates how to adapt large language models to mimic an individual author's style using only a few example abstracts, a task made harder by the formal nature of scientific writing. It proposes three style‑conditioning methods—contrastive activation steering, a network predicting steering vectors, and a hypernetwork predicting LoRA adapters—and finds that while fine‑tuning captures the strongest style signal, it harms fluency; the hypernetwork offers the best balance between style imitation and output quality for both seen and unseen authors. The authors also show that steering can be performed at the author level by contrasting author abstracts against style‑neutral generations for the same content, eliminating the need for a predefined style inventory and outperforming inventory‑based approaches.
whyItMatters":"The study provides practical techniques for author‑style transfer in scientific writing, revealing a trade‑off between style fidelity and fluency and demonstrating that hypernetworks can effectively balance these aspects."
By Leonard Popp, Danni Liu, Supriti Sinhamahapatra, Jan Niehues
The paper investigates how post‑training of language models introduces or amplifies an AI‑like writing style and whether this style has a localized internal signature. By comparing human text, base‑model outputs, and aligned‑model outputs with matched human‑source prefixes, the authors find that aligned generations are less similar to human corpora and more easily detected as AI‑like. They propose PASTA, a training‑free method that estimates a post‑training alignment signature from residual contrasts and ablates the corresponding direction during decoding, reducing detection rates across multiple models and detectors while preserving relevance and coherence.
By Aniket Anand, Janvijay Singh, Zhewei Sun, Dilek Hakkani-T\"ur, Nick Feamster
arXiv:2605.30526v2 Announce Type: replace
Abstract: Aligned language models often exhibit a recognizable AI-like style, yet its connection to post-training and internal representations remains poorly...
By Aniket Anand, Janvijay Singh, Zhewei Sun, Dilek Hakkani-T\"ur, Nick Feamster
The paper investigates whether activation steering can capture the complex, multidimensional nature of authorship style. By using structured contrastive prompting along rhetorically motivated dimensions, the authors construct rich style representations directly in activation space, revealing a shared authorship backbone with aspect‑specific residuals. They propose Aspect‑Aware Activation Steering (A3S), a training‑free framework that merges per‑aspect contrastive directions, employs interference‑aware aggregation, and tunes steering strength per instance, achieving better authorship style transfer and outperforming a trained baseline on out‑of‑domain benchmarks.
By Hieu Tran, Calvin Bao, Marine Carpuat
The paper introduces a new authorship attribution model that represents each author with multiple consistent but distinct writing styles, rather than a single style profile. Experiments across four scenarios—audience size, temporal distance, preceding context, and genre norms—show that authors vary significantly in how idiolectic their writing is, and that certain settings yield text that is more informative about authorship. The authors demonstrate that modeling multiple styles per author leads to significant performance improvements over traditional single-style approaches.
By Omkar Yadav, Michael Jiang, Kenan Alkiek, Jian Zhu, David Jurgens
The study investigates bias in large language model (LLM) judges by having ten LLMs evaluate narrative constraint selections rather than generated text. Results show that self-preference largely disappears under blind evaluation when quality and evaluator severity are controlled, but self- and other-labels alone shift scores bidirectionally when quality is matched. The authors conclude that authorship attribution drives evaluation bias and that open-ended, ground‑truth‑free tasks can effectively study LLM judge behavior.
By Songeun Chae, Min Kim, Donghoon Jung, Seojin Choi, Seohyon Jung
The paper "Evaluating Style-Personalized Text Generation: Challenges and Directions" examines the difficulties of assessing text that is tailored to individual users’ styles. It critiques common metrics such as BLEU, embeddings, and LLM-as-judges, and introduces a style discrimination benchmark covering domain discrimination, authorship attribution, and LLM-generated personalized versus non-personalized discrimination across eight writing tasks. The study finds that ensembles of diverse evaluation metrics outperform single-evaluator approaches and offers guidance for reliable assessment of style-personalized generation.
By Anubhav Jangra, Bahareh Sarrafzadeh, Silviu Cucerzan, Adrian de Wynter, Sujay Kumar Jauhar
The study examines whether large language models (LLMs) show syntactic convergence with human users by analyzing turn‑adjacent reuse of context‑free grammar (CFG) rules in 1,901 matched positions across sixteen open‑weight Llama and Gemma models. All models exhibited greater CFG‑rule overlap with the preceding human turn than with unrelated primes, especially for lower‑frequency rules, and instruction‑tuned models showed even higher overlap with the actual prime than the human responses they replaced. However, instruction tuning also increased overlap with unrelated primes, reduced the actual‑versus‑random difference, and lowered conditional rule‑reuse odds when controlling for rule‑set size, while consistently producing responses with higher lexical and semantic similarity to the preceding turn.
By Zandi Eberstadt
arXiv:2310.00436v2 Announce Type: replace
Abstract: Authorship identification uses patterns in writing to infer who wrote a text, but those patterns also reflect topic, genre, and register. This surv...
By Haining Wang
The paper investigates whether language models possess latent rhetorical abilities, specifically analogy reasoning, before instruction tuning. By extracting an "Analogy Vector" from the residual stream, the authors demonstrate that this vector can steer explanation style across several instruction-tuned models, improving performance metrics. Their findings reveal a continuous gradient of behavioral accessibility tied to the amount of instruction exposure during training, challenging the binary base‑vs‑instruct distinction and suggesting that instruction exposure shapes rhetorical knowledge in activation space.
By Shengjie Xu
arXiv:2608. 01666v2 Announce Type: replace-cross Abstract: However, whether these judges truly evaluate the scientific substance of ideas or are influenced by superficial stylistic presentation remains an open question.
By Fengxian Ji, Yuke Li, Jingpu Yang, Juanfan Wu, Fan Zhang, Zhexuan Cui, Yu Xie, Min Peng, Qianqian Xie, Xiuying Chen, Zhuohan Xie
The paper introduces GLASS, a training‑free framework that personalizes text generation by combining global user‑style priors extracted with sparse autoencoders and local contrastive style vectors built from clustered interaction scenarios. During inference, GLASS injects these global and local vectors into different model layers, achieving context‑aware personalization without retrieval or parameter updates. Experiments on LaMP and LongLaMP datasets demonstrate that GLASS surpasses retrieval, fine‑tuning, and other steering baselines on ROUGE metrics and LLM‑as‑judge evaluations, and its SAE‑based representations show robustness to topic and length shifts, indicating better disentanglement of style from content.
By Liuji Chen, Zeyu Zhang, Xinyuan Zhang, Shuai Nie, Qiang Liu, Shu Wu, Liang Wang