mlsanityData to 7 Oct 2026

StylePTB: A Compositional Benchmark for Fine-grained Controllable Text Style Transfer

arXiv 2104.05196 ↗First posted 12 Apr 2021cs.CL cs.AI cs.LGcited by 34

Not on the map. Why?

Abstract (arXiv)

Text style transfer aims to controllably generate text with targeted stylistic changes while maintaining core meaning from the source sentence constant. Many of the existing style transfer benchmarks primarily focus on individual high-level semantic changes (e.g. positive to negative), which enable controllability at a high level but do not offer fine-grained control involving sentence structure, emphasis, and content of the sentence. In this paper, we introduce a large-scale benchmark, StylePTB, with (1) paired sentences undergoing 21 fine-grained stylistic changes spanning atomic lexical, syntactic, semantic, and thematic transfers of text, as well as (2) compositions of multiple transfers which allow modeling of fine-grained stylistic changes as building blocks for more complex, high-level transfers. By benchmarking existing methods on StylePTB, we find that they struggle to model fine-grained changes and have an even more difficult time composing multiple styles. As a result, StylePTB brings novel challenges that we hope will encourage future research in controllable text style transfer, compositional models, and learning disentangled representations. Solving these challenges would present important steps towards controllable text generation.

Key facts. Cited by 34 papers in the map (data through 7 Oct 2026). Builds on 31 of its 60 references that are in the map. First posted on arXiv 12 Apr 2021.

Builds on

31 of 60 references.

References · 31

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  2. 2014Sequence to Sequence Learning with Neural Networkscited by 2,974
  3. 2016Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotationscited by 2,646
  4. 2019The Woman Worked as a Babysitter: On Biases in Language Generationcited by 540
  5. 2020Null It Out: Guarding Protected Attributes by Iterative Nullspace Projectioncited by 442
  6. 2017Toward Controlled Generation of Textcited by 271
  7. 2017Style Transfer from Non-Parallel Text by Cross-Alignmentcited by 246
  8. 2018Dear Sir or Madam, May I introduce the GYAFC Dataset: Corpus, Benchmarks and Metrics for Formality Style Transfercited by 209
  9. 2017Survey of the State of the Art in Natural Language Generation: Core tasks, applications and evaluationcited by 206
  10. 2018Delete, Retrieve, Generate: A Simple Approach to Sentiment and Style Transfercited by 191
  11. 2020Reformulating Unsupervised Style Transfer as Paraphrase Generationcited by 181
  12. 2020Towards Debiasing Sentence Representationscited by 170
19 more
  1. 2017Style Transfer in Text: Exploration and Evaluationcited by 145
  2. 2015Character-Aware Neural Language Modelscited by 129
  3. 2018Disentangled Representation Learning for Non-Parallel Text Style Transfercited by 127
  4. 2018Style Transfer Through Back-Translationcited by 124
  5. 2020Towards Controllable Biases in Language Generationcited by 112
  6. 2017Generating Sentences by Editing Prototypescited by 88
  7. 2019Transforming Delete, Retrieve, Generate Approach for Controlled Text Style Transfercited by 81
  8. 2020Politeness Transfer: A Tag and Generate Approachcited by 78
  9. 2018A Retrieve-and-Edit Framework for Predicting Structured Outputscited by 73
  10. 2020A Probabilistic Formulation of Unsupervised Text Style Transfercited by 70
  11. 2017Relevance of Unsupervised Metrics in Task-Oriented Dialogue for Evaluating Natural Language Generationcited by 61
  12. 2019Evaluating Style Transfer for Textcited by 59
  13. 2017Shakespearizing Modern Language Using Copy-Enriched Sequence-to-Sequence Modelscited by 57
  14. 2019"Mask and Infill" : Applying Masked Language Model to Sentiment Transfercited by 48
  15. 2020Hooks in the Headline: Learning to Generate Headlines with Controlled Stylescited by 46
  16. 2020On Learning Text Style Transfer with Direct Rewardscited by 24
  17. 2019Adapting Language Models for Non-Parallel Author-Stylized Rewritingcited by 23
  18. 2020Exploring Contextual Word-level Style Relevance for Unsupervised Style Transfercited by 15
  19. 2020On Emergent Communication in Competitive Multi-Agent Teamscited by 8

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 5; 2022: 12; 2023: 10; 2024: 3; 2025: 1; 2026: 3.

34 papers.

Sort:

By area:

  1. 2022Extracting Latent Steering Vectors from Pretrained Language Modelscited by 217
  2. 2021ExT5: Towards Extreme Multi-Task Scaling for Transfer Learningcited by 161
  3. 2024Benchmarking Complex Instruction-Following with Multiple Constraints Compositioncited by 147
  4. 2022Differentially Private Language Models for Secure Data Sharingcited by 68
  5. 2021Tailor: Generating and Perturbing Text with Semantic Controlscited by 67
  6. 2021Sequence-to-Sequence Learning with Latent Neural Grammarscited by 30
  7. 2021Evaluating the Evaluation Metrics for Style Transfer: A Case Study in Multilingual Formality Transfercited by 23
  8. 2022Tailor: A Prompt-Based Approach to Attribute-Based Controlled Text Generationcited by 22
  9. 2022Why is constrained neural language generation particularly challenging?cited by 19
  10. 2023Med-EASi: Finely Annotated Dataset and Models for Controllable Simplification of Medical Textscited by 16
  11. 2023A Call for Standardization and Validation of Text Style Transfer Evaluationcited by 13
  12. 2022Human Judgement as a Compass to Navigate Automatic Metrics for Formality Transfercited by 11
22 more
  1. 2022An Extensible Plug-and-Play Method for Multi-Aspect Controllable Text Generationcited by 10
  2. 2022Towards Robust and Semantically Organised Latent Representations for Unsupervised Text Style Transfercited by 7
  3. 2023Extrapolative Controlled Sequence Generation via Iterative Refinementcited by 7
  4. 2023Fine-grained Text Style Transfer with Diffusion-Based Language Modelscited by 7
  5. 2023Seen to Unseen: Exploring Compositional Generalization of Multi-Attribute Controllable Dialogue Generationcited by 7
  6. 2022Compositional Generalisation with Structured Reordering and Fertility Layerscited by 6
  7. 2022Nano: Nested Human-in-the-Loop Reward Learning for Few-shot Language Model Controlcited by 6
  8. 2023XATU: A Fine-grained Instruction-based Benchmark for Explainable Text Updatescited by 6
  9. 2021From Theories on Styles to their Transfer in Text: Bridging the Gap with a Hierarchical Surveycited by 5
  10. 2022Brainish: Formalizing A Multimodal Language for Intelligence and Consciousnesscited by 4
  11. 2024Strengthening Structural Inductive Biases by Pre-training to Perform Syntactic Transformationscited by 4
  12. 2022Few-Shot Table-to-Text Generation with Prefix-Controlled Generatorcited by 3
  13. 2023Adapter-TST: A Parameter Efficient Method for Multiple-Attribute Text Style Transfercited by 2
  14. 2023Towards Zero-Shot Functional Compositionality of Language Modelscited by 1
  15. 2023Improving Grammar-based Sequence-to-Sequence Modeling with Decomposition and Constraintscited by 1
  16. 2022On Text Style Transfer via Style Masked Language Modelscited by 0
  17. 2023Learning a Formality-Aware Japanese Sentence Representationcited by 0
  18. 2024GCOF: Self-iterative Text Generation for Copywriting Using Large Language Modelcited by 0
  19. 2025Challenging Assumptions in Learning Generic Text Style Embeddingscited by 0
  20. 2026The Significance of Style Diversity in Annotation-Free Synthetic Data Generationcited by 0
  21. 2026STEB: Style Text Embedding Benchmarkcited by 0
  22. 2026Do Recipes Have Personas? Characterizing and Generating Creator Style in Attributed Procedural Graphscited by 0