Quantizers can be ranked by privacy using r_Q, a rate constant built from the loss gap and variance of low-loss quantized checkpoints along the training trajectory.
Hardware-aware dnn compression via diverse pruning and mixed-precision quantization
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ML 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Membership Inference Risks in Quantized Models: A Theoretical and Empirical Study
Quantizers can be ranked by privacy using r_Q, a rate constant built from the loss gap and variance of low-loss quantized checkpoints along the training trajectory.