DPS-MOZO samples each generated token from the product of per-example distributions mixed with the zero-shot distribution to make in-context learning differentially private without additive noise.
On the privacy risk of in-context learning
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Differentially Private In-context Learning via Sampling Few-shot Mixed with Zero-shot Outputs
DPS-MOZO samples each generated token from the product of per-example distributions mixed with the zero-shot distribution to make in-context learning differentially private without additive noise.