Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

RADAR: Run-time Adversarial Weight Attack Detection and Accuracy Recovery

About

Adversarial attacks on Neural Network weights, such as the progressive bit-flip attack (PBFA), can cause a catastrophic degradation in accuracy by flipping a very small number of bits. Furthermore, PBFA can be conducted at run time on the weights stored in DRAM main memory. In this work, we propose RADAR, a Run-time adversarial weight Attack Detection and Accuracy Recovery scheme to protect DNN weights against PBFA. We organize weights that are interspersed in a layer into groups and employ a checksum-based algorithm on weights to derive a 2-bit signature for each group. At run time, the 2-bit signature is computed and compared with the securely stored golden signature to detect the bit-flip attacks in a group. After successful detection, we zero out all the weights in a group to mitigate the accuracy drop caused by malicious bit-flips. The proposed scheme is embedded in the inference computation stage. For the ResNet-18 ImageNet model, our method can detect 9.6 bit-flips out of 10 on average. For this model, the proposed accuracy recovery scheme can restore the accuracy from below 1% caused by 10 bit flips to above 69%. The proposed method has extremely low time and storage overhead. System-level simulation on gem5 shows that RADAR only adds <1% to the inference time, making this scheme highly suitable for run-time attack detection and mitigation.

Jingtao Li, Adnan Siraj Rakin, Zhezhi He, Deliang Fan, Chaitali Chakrabarti• 2021

Related benchmarks

TaskDatasetResultRank
White-box robustness against single point failure attacksWikiText-2
Original Perplexity (PPL)5.03
8
LLM InferenceOPT-125M
Latency (ms)76.2
4
LLM InferenceQwen2.5 0.5B
Latency (ms)238.4
4
LLM InferenceLLaMA2-7B
Latency (ms)1.94e+3
4
Showing 4 of 4 rows

Other info

Follow for update