arXiv Machine Learning October 3, 2026 By Ioana Marinescu, Eric Karl Oermann, Kyunghyun Cho

Reason in Style: Discovering and Controlling Style in Language Models

Read the original on arXiv Machine Learning →

The paper introduces an unsupervised algorithm that separates content and style in language model outputs, enabling explicit style control. Applied to over 100K traces from nine teacher models, the method uncovers six recurring, imbalanced styles and fine‑tunes student models to follow them. This style conditioning improves Pass@$k$ on six math reasoning benchmarks and shows that style influences problem‑solving success, offering a new avenue for controlling and enhancing model performance.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Machine Learning.

arXiv Computation and Language

Sep 16

Style-Debiased DPO: Updating LLM Knowledge with Factuality-Aware Synthetic Preference Data

The paper introduces Style‑Debiased DPO (SD‑DPO), a method that refines large language models’ ability to retrieve stored knowledge by using preference optimization that corrects for style differences while preserving factual accuracy. SD‑DPO evaluates on the EntiGraph storing‑side framework and outperforms baseline CPT on the QuALITY reading‑comprehension benchmark, achieving higher accuracy with far fewer training tokens. In a knowledge‑editing setting (AToKE), SD‑DPO attains an overall accuracy of 0.982, correctly answering queries with either new or old facts based on the requested time period.

By Takayuki Yamamoto, Daisuke Kawahara

llms reinforcement-learning benchmarks

arXiv AI

Jun 19

Too long; didn't solve

arXiv:2604. 07593v2 Announce Type: replace Abstract: Mathematical benchmarks consisting of a range of mathematics problems are widely used to evaluate the reasoning abilities of large language models, yet little is known about how their structural properties influence model behaviour.

By Luc\'ia M. Cabrera, Isaac Saxton-Knight, Jocelyn D'Arcy

llms benchmarks safety

arXiv Computation and Language

Sep 23

ReasonLab: A Controlled and Auditable Evaluation of Prompting Techniques for Multiple-Choice QA

arXiv:2607.14109v2 Announce Type: replace Abstract: Probing the capabilities of Large Language Models (LLMs) and building robust solutions for Multiple-Choice Question Answering (MCQA) remain central...

By Inder Preet, Shuxin Lin, Dhaval Patel

llms nlp benchmarks

arXiv Computation and Language

Oct 5

RT-SFT: Text Style Transfer from Non-Parallel Corpora by Roundtrip Translation

RT‑SFT is a method for text style transfer that uses roundtrip translation through a pivot language to strip stylistic information from monolingual corpora, creating pseudo‑parallel data. This data is then used to LoRA‑finetune an instruction‑tuned large language model as a stylizer, allowing the model to rewrite sentences in a target style while preserving meaning. Experiments across four style domains show that RT‑SFT surpasses state‑of‑the‑art approaches, including few‑shot in‑context learning, and offers effective retrieval augmentation for expert style domains with strict terminology.

By Ruoxi Liu, Philipp Koehn

llms fine-tuning benchmarks

arXiv AI

Jul 17

Simplicity Paradox: Debunking myths about prompting and datasets for LLM evaluation

arXiv:2607. 14109v1 Announce Type: cross Abstract: Probing the capabilities of Large Language Models (LLMs) and building robust solutions for Multiple-Choice Question Answering (MCQA) remain central challenges in natural language understanding.

By Inder Preet, Shuxin Lin, Dhaval Patel

llms nlp benchmarks

arXiv AI

Jun 30

Beyond Scaling Law: A Data-Efficient Distillation Framework for Reasoning

arXiv:2508. 09883v2 Announce Type: replace-cross Abstract: Large language models (LLMs) demonstrate remarkable reasoning capabilities in tasks such as algorithmic coding and mathematical problem-solving.

By Xiaojun Wu, Xiaoguang Jiang, Huiyang Li, Jucai Zhai, Dengfeng Liu, Qiaobo Hao, Huang Liu, Zhiguo Yang, Ji Xie, Ninglun Gu, Jin Yang, Kailai Zhang, Yelun Bao, Jun Wang

llms reinforcement-learning fine-tuning efficiency benchmarks

One email a morning, machine-written

Related stories

Style-Debiased DPO: Updating LLM Knowledge with Factuality-Aware Synthetic Preference Data

Too long; didn't solve

ReasonLab: A Controlled and Auditable Evaluation of Prompting Techniques for Multiple-Choice QA

RT-SFT: Text Style Transfer from Non-Parallel Corpora by Roundtrip Translation

Simplicity Paradox: Debunking myths about prompting and datasets for LLM evaluation

Beyond Scaling Law: A Data-Efficient Distillation Framework for Reasoning