Evidence · DiggingBeagle record
SURA 10-image trusted probes suppress federated-learning backdoors unless attacker knows exact repair set
Original SURA paper screens federated client updates with small clean probe sets and tests an adversary knowing that very set. One adaptive setting raises backdoor attack success to 94.91%, highlighting reference-set secrecy and limited transfer to other architectures.
- Source role
- primary disclosure
Evidence record
Original SURA paper screens federated client updates with small clean probe sets and tests an adversary knowing that very set. One adaptive setting raises backdoor attack success to 94.91%, highlighting reference-set secrecy and limited transfer to other architectures.
Claim-level citations (1)
- supportsSURA backdoor screening failed when attackers learned its ten clean reference images: In SURA’s tested adaptive condition, knowing the exact ten-image trusted probe set raised the reported backdoor attack-success rate to 94.91%.
Section 3.6, adaptive attacker with known trusted set