# Learning Interpretable Style Embeddings via Prompting LLMs Source: [Learning Interpretable Style Embeddings via Prompting LLMs]() Collection index: [llms.txt]() Data updated: 2025-10-26T01:49:52.893Z Authors: Ajay Patel, Delip Rao, Ansh Kothary, Kathleen McKeown, Chris Callison-Burch Published: 2023-05-22T04:07:54.000Z ## Abstract Style representation learning builds content-independent representations of author style in text. Stylometry, the analysis of style in text, is often performed by expert forensic linguists and no large dataset of stylometric annotations exists for training. Current style representation learning uses neural methods to disentangle style from content to create style vectors, however, these approaches result in uninterpretable representations, complicating their usage in downstream applications like authorship attribution where auditing and explainability is critical. In this work, we use prompting to perform stylometry on a large number of texts to create a synthetic dataset and train human-interpretable style representations we call LISA embeddings. We release our synthetic stylometry dataset and our interpretable style models as resources. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Authorship Verification on PAN AV 2015 (test)](): Authorship Verification; PAN AV 2015 (test); ROC-AUC (Greek): 0.48. - [Authorship Verification on PAN AV 2013 (test)](): Authorship Verification; PAN AV 2013 (test); ROC-AUC (Greek): 0.51. - [Style Representation Evaluation on STEL-or-Content Multilingual (averaged across languages)](): Style Representation Evaluation; STEL-or-Content Multilingual (averaged across languages); Simplicity Score: 15. - [Style Representation Evaluation on STEL-or-Content Cross-lingual (averaged across languages)](): Style Representation Evaluation; STEL-or-Content Cross-lingual (averaged across languages); Formality: 0.27. - [Authorship Verification on PAN AV 2014 (test)](): Authorship Verification; PAN AV 2014 (test); ROC-AUC (Greek): 0.46. - [Authorship Verification on PAN Average 2013-2015 (test)](): Authorship Verification; PAN Average 2013-2015 (test); Greek Avg ROC-AUC: 0.48. ## Original sources and implementation [Original paper]() [PDF]()