ContextLeak is the first empirical framework to audit worst-case information leakage in private in-context learning by inserting identifiable canary tokens and measuring their presence in model outputs.
Sergey Ioffe and Christian Szegedy
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
roles
background 1polarities
unclear 1representative citing papers
DPQuant uses epoch-wise probabilistic layer rotation and DP loss sensitivity to quantize only a changing subset of layers, reducing accuracy degradation from quantization noise in DP-SGD and delivering up to 2.21x throughput gains with under 2% accuracy drop.
Memorization in language models increases log-linearly with model capacity, data duplication count, and prompt context length.
A survey of differential privacy theory, mechanisms, applications, and user-facing issues.
citing papers explorer
-
ContextLeak: Auditing Leakage in Private In-Context Learning Methods
ContextLeak is the first empirical framework to audit worst-case information leakage in private in-context learning by inserting identifiable canary tokens and measuring their presence in model outputs.
-
DPQuant: Efficient and Differentially-Private Model Training via Dynamic Quantization Scheduling
DPQuant uses epoch-wise probabilistic layer rotation and DP loss sensitivity to quantize only a changing subset of layers, reducing accuracy degradation from quantization noise in DP-SGD and delivering up to 2.21x throughput gains with under 2% accuracy drop.
-
Quantifying Memorization Across Neural Language Models
Memorization in language models increases log-linearly with model capacity, data duplication count, and prompt context length.
-
A Comprehensive Guide to Differential Privacy: From Theory to User Expectations
A survey of differential privacy theory, mechanisms, applications, and user-facing issues.