Dynamic quantization creates side channels allowing partial or full recovery of other users' batched data in at least four popular ML frameworks.
Architectural backdoors for within-batch data stealing and model inference manipulation
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CR 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Quantamination: Dynamic Quantization Leaks Your Data Across the Batch
Dynamic quantization creates side channels allowing partial or full recovery of other users' batched data in at least four popular ML frameworks.