PersonaLedger LLM simulator achieves AUC 0.70 for fraud detection at epsilon=1 from DP inputs but shows significant distribution drift due to learned priors overriding input statistics on temporal and demographic features.
Winning the nist contest: A scalable and general approach to differentially private synthetic data.Journal of Privacy and Confidentiality, 11(3),
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Evaluating LLM Simulators as Differentially Private Data Generators
PersonaLedger LLM simulator achieves AUC 0.70 for fraud detection at epsilon=1 from DP inputs but shows significant distribution drift due to learned priors overriding input statistics on temporal and demographic features.