Controlled Generation for Private Synthetic Text
Not on the map. Why?
Abstract (arXiv)
Text anonymization is essential for responsibly developing and deploying AI in high-stakes domains such as healthcare, social services, and law. In this work, we propose a novel methodology for privacy-preserving synthetic text generation that leverages the principles of de-identification and the Hiding In Plain Sight (HIPS) theory. Our approach introduces entity-aware control codes to guide controllable generation using either in-context learning (ICL) or prefix tuning. The ICL variant ensures privacy levels consistent with the underlying de-identification system, while the prefix tuning variant incorporates a custom masking strategy and loss function to support scalable, high-quality generation. Experiments on legal and clinical datasets demonstrate that our method achieves a strong balance between privacy protection and utility, offering a practical and effective solution for synthetic text generation in sensitive domains.
Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 28 of its 44 references that are in the map. First posted on arXiv 30 Sep 2025.
Builds on
28 of 44 references.
References · 28
- 2021Prefix-Tuning: Optimizing Continuous Prompts for Generationcited by 3,915
- 2016Deep Learning with Differential Privacycited by 2,728
- 2020Extracting Training Data from Large Language Modelscited by 2,189
- 2019CTRL: A Conditional Transformer Language Model for Controllable Generationcited by 762
- 2021MAUVE: Measuring the Gap Between Neural Text and Human Text using Divergence Frontierscited by 470
- 2023TinyStories: How Small Can Language Models Be and Still Speak Coherent English?cited by 449
- 2023Sheared LLaMA: Accelerating Language Model Pre-training via Structured Pruningcited by 415
- 2021Large Language Models Can Be Strong Differentially Private Learnerscited by 412
- 2021Differentially Private Fine-tuning of Language Modelscited by 400
- 2019Analysing Mathematical Reasoning Abilities of Neural Modelscited by 335
- 2018Differentially Private Generative Adversarial Networkcited by 235
- 2022ZeroGen: Efficient Zero-shot Learning via Dataset Generationcited by 211
16 more
- 2022Is GPT-3 a Good Data Annotator?cited by 208
- 2023Large Language Models for Code: Security Hardening and Adversarial Testingcited by 155
- 2020How Does NLP Benefit Legal System: A Summary of Legal Artificial Intelligencecited by 129
- 2019Privacy- and Utility-Preserving Textual Analysis via Calibrated Multivariate Perturbationscited by 119
- 2020GS-WGAN: A Gradient-Sanitized Approach for Learning Differentially Private Generatorscited by 117
- 2022Synthetic Text Generation with Differential Privacy: A Simple and Practical Recipecited by 108
- 2022Differentially Private Language Models for Secure Data Sharingcited by 68
- 2023Harnessing large-language models to generate private synthetic textcited by 65
- 2022The Limits of Word Level Differential Privacycited by 63
- 2023Locally Differentially Private Document Generation Using Zero Shot Promptingcited by 63
- 2019Leveraging Hierarchical Representations for Preserving Privacy and Utility in Textcited by 50
- 2021ADePT: Auto-encoder based Differentially Private Text Transformationcited by 31
- 2024Evaluating Differentially Private Synthetic Data Generation in High-Stakes Domainscited by 12
- 2024On the Vulnerability of Text Sanitizationcited by 9
- 2025SynthTextEval: Synthetic Text Data Generation and Evaluation for High-Stakes Domainscited by 7
- 2022Mention Annotations Alone Enable Efficient Domain Adaptation for Coreference Resolutioncited by 6
Built on by
Citing papers per year: 2026: 2.
2 papers.