SynBench benchmarks DP text generators across nine datasets and uses a new MIA to show that public pre-training on portions of private data overestimates synthetic text quality and breaks DP privacy bounds.
Matthieu Meeus, Lukas Wutschitz, Santiago Zanella-B´eguelin, Shruti Tople, and Reza Shokri
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
representative citing papers
LLMs engage in spontaneous persuasion in virtually all multi-turn conversations by favoring information-based strategies like logic and evidence, in contrast to human responses that rely more on social influence and negative emotions.
citing papers explorer
-
SynBench: A Benchmark for Differentially Private Text Generation
SynBench benchmarks DP text generators across nine datasets and uses a new MIA to show that public pre-training on portions of private data overestimates synthetic text quality and breaks DP privacy bounds.
-
Spontaneous Persuasion: An Audit of Model Persuasiveness in Everyday Conversations
LLMs engage in spontaneous persuasion in virtually all multi-turn conversations by favoring information-based strategies like logic and evidence, in contrast to human responses that rely more on social influence and negative emotions.