ContinuousBench: Can Differentially Private Synthetic Text Improve Capabilities? 事件

ContinuousBench: Can Differentially Private Synthetic Text Improve Capabilities? arXiv:2606.01849v1 Announce Type: cross Abstract: Differentially private (DP) text synthesis promises to unlock sensitive corpora for model training, but it remains unclear whether DP synthetic data transmits genuinely new knowledge and capabilities present only in those corpora. This is because existing evaluations rely on tasks that are nearly solvable without training, so strong benchmark performance does not es