ACTG-ARL: Differentially Private Conditional Text Generation with RL-Boosted Control

TL;DR

This paper introduces ACTG-ARL, a hierarchical framework with RL-boosted control for differentially private synthetic text generation, significantly improving quality and control while maintaining strong privacy guarantees.

Contribution

The paper presents a novel hierarchical approach for DP text synthesis and introduces Anchored RL to enhance conditional control, achieving notable quality improvements.

Findings

+20% MAUVE score over prior work

Effective integration of feature learning and conditional generation

Anchored RL improves instruction-following in DP text synthesis

Abstract

Generating high-quality synthetic text under differential privacy (DP) is critical for training and evaluating language models without compromising user privacy. Prior work on synthesizing DP datasets often fail to preserve key statistical attributes, suffer utility loss from the noise required by DP, and lack fine-grained control over generation. To address these challenges, we make two contributions. First, we introduce a hierarchical framework that decomposes DP synthetic text generation into two subtasks: feature learning and conditional text generation. This design explicitly incorporates learned features into the generation process and simplifies the end-to-end synthesis task. Through systematic ablations, we identify the most effective configuration: a rich tabular schema as feature, a DP tabular synthesizer, and a DP fine-tuned conditional generator, which we term ACTG (Attribute-Conditioned Text Generation). Second, we propose Anchored RL (ARL), a post-training method that improves the instruction-following ability of ACTG for conditional generation. ARL combines RL to boost control with an SFT anchor on best-of-$N$ data to prevent reward hacking. Together, these components form our end-to-end algorithm ACTG-ARL, which advances both the quality of DP synthetic text (+20% MAUVE over prior work) and the control of the conditional generator under strong privacy guarantees.