mlsanityData to 14 Oct 2026

Reformulating Unsupervised Style Transfer as Paraphrase Generation

arXiv 2010.05700 ↗First posted 12 Oct 2020cs.CLcited by 183

Not on the map. Why?

Cited by 24 in LLM safety and security, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

Modern NLP defines the task of style transfer as modifying the style of a given sentence without appreciably changing its semantics, which implies that the outputs of style transfer systems should be paraphrases of their inputs. However, many existing systems purportedly designed for style transfer inherently warp the input's meaning through attribute transfer, which changes semantic properties such as sentiment. In this paper, we reformulate unsupervised style transfer as a paraphrase generation problem, and present a simple methodology based on fine-tuning pretrained language models on automatically generated paraphrase data. Despite its simplicity, our method significantly outperforms state-of-the-art style transfer systems on both human and automatic evaluations. We also survey 23 style transfer papers and discover that existing automatic metrics can be easily gamed and propose fixed variants. Finally, we pivot to a more real-world style transfer setting by collecting a large dataset of 15M sentences in 11 diverse styles, which we use for an in-depth analysis of our system.

Key facts. Cited by 183 papers in the map (data through 14 Oct 2026). First posted on arXiv 12 Oct 2020.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 38; 2022: 44; 2023: 34; 2024: 33; 2025: 20; 2026: 14.

183 papers.

Sort:

By area:

  1. 2023Paraphrasing evades detectors of AI-generated text, but retrieval is an effective defensecited by 429
  2. 2022RLPrompt: Optimizing Discrete Text Prompts with Reinforcement Learningcited by 366
  3. 2022ROSCOE: A Suite of Metrics for Scoring Step-by-Step Reasoningcited by 177
  4. 2021Mind the Style of Text! Adversarial and Backdoor Attacks Based on Text Style Transfercited by 158
  5. 2024Benchmarking Complex Instruction-Following with Multiple Constraints Compositioncited by 152
  6. 2022Pretrained Language Models for Text Generation: A Surveycited by 122
  7. 2021A Recipe For Arbitrary Text Style Transfer with Large Language Modelscited by 107
  8. 2021The Perils of Using Mechanical Turk to Evaluate Open-Ended Text Generationcited by 97
  9. 2024Large Concept Models: Language Modeling in a Sentence Representation Spacecited by 89
  10. 2021NL-Augmenter: A Framework for Task-Sensitive Natural Language Augmentationcited by 73
  11. 2021Controlled Text Generation as Continuous Optimization with Multiple Constraintscited by 70
  12. 2022Mix and Match: Learning-free Controllable Text Generation using Energy Language Modelscited by 66
38 more
  1. 2021Compression, Transduction, and Creation: A Unified Framework for Evaluating Natural Language Generationcited by 60
  2. 2022LENS: A Learnable Evaluation Metric for Text Simplificationcited by 54
  3. 2022BITE: Textual Backdoor Attacks with Iterative Trigger Injectioncited by 53
  4. 2022Multi-VALUE: A Framework for Cross-Dialectal English NLPcited by 53
  5. 2021XFORMAL: A Benchmark for Multilingual Formality Style Transfercited by 52
  6. 2021Text Detoxification using Large Pre-trained Neural Modelscited by 52
  7. 2022Prompt-and-Rerank: A Method for Zero-Shot and Few-Shot Arbitrary Textual Style Transfer with Small Language Modelscited by 49
  8. 2022VALUE: Understanding Dialect Disparity in NLUcited by 47
  9. 2021Phrase-BERT: Improved Phrase Embeddings from BERT with an Application to Corpus Explorationcited by 44
  10. 2024Detection and Measurement of Syntactic Templates in Generated Textcited by 44
  11. 2023ChatGPT to Replace Crowdsourcing of Paraphrases for Intent Classification: Higher Diversity and Comparable Model Robustnesscited by 41
  12. 2021LEWIS: Levenshtein Editing for Unsupervised Text Style Transfercited by 40
  13. 2022Human-Machine Collaboration Approaches to Build a Dialogue Dataset for Hate Speech Counteringcited by 40
  14. 2023You Only Prompt Once: On the Capabilities of Prompt Learning on Large Language Models to Tackle Toxic Contentcited by 39
  15. 2023Backdoor Attacks and Countermeasures in Natural Language Processing Models: A Comprehensive Security Reviewcited by 39
  16. 2022Detoxifying Text with MaRCo: Controllable Revision with Experts and Anti-Expertscited by 36
  17. 2021StylePTB: A Compositional Benchmark for Fine-grained Controllable Text Style Transfercited by 34
  18. 2024Aligning Language Models with Demonstrated Feedbackcited by 34
  19. 2022UCTopic: Unsupervised Contrastive Learning for Phrase Representations and Topic Miningcited by 33
  20. 2023Attention-Enhancing Backdoor Attacks Against BERT-based Modelscited by 32
  21. 2023ChatGPT as an Attack Tool: Stealthy Textual Backdoor Attack via Blackbox Generative Model Triggercited by 31
  22. 2023TrojLLM: A Black-box Trojan Prompt Attack on Large Language Modelscited by 31
  23. 2022arXivEdits: Understanding the Human Revision Process in Scientific Writingcited by 26
  24. 2021ConvFiT: Conversational Fine-Tuning of Pretrained Language Modelscited by 25
  25. 2021Evaluating the Evaluation Metrics for Style Transfer: A Case Study in Multilingual Formality Transfercited by 25
  26. 2023Don't lose the message while paraphrasing: A study on content preserving style transfercited by 25
  27. 2022CORE: A Retrieve-then-Edit Framework for Counterfactual Data Generationcited by 24
  28. 2023Assessing the Quality of Multiple-Choice Questions Using GPT-4 and Rule-Based Methodscited by 24
  29. 2021CrossAug: A Contrastive Data Augmentation Method for Debiasing Fact Verification Modelscited by 23
  30. 2021Paraphrastic Representations at Scalecited by 21
  31. 2021Multi-granularity Textual Adversarial Attack with Behavior Cloningcited by 21
  32. 2021Few-shot Controllable Style Transfer for Low-Resource Multilingual Settingscited by 21
  33. 2021A Review of Human Evaluation for Style Transfercited by 19
  34. 2023Setting the Trap: Capturing and Defeating Backdoors in Pretrained Language Models through Honeypotscited by 19
  35. 2023ParaGuide: Guided Diffusion Paraphrasers for Plug-and-Play Textual Style Transfercited by 18
  36. 2024StyleRemix: Interpretable Authorship Obfuscation via Distillation and Perturbation of Style Elementscited by 18
  37. 2025BadJudge: Backdoor Vulnerabilities of LLM-as-a-Judgecited by 18
  38. 2021Does It Capture STEL? A Modular, Similarity-based Linguistic Style Evaluation Frameworkcited by 17

50 most cited of 183 All 183 (JSON)