.png){kind=icon}
.png){kind=icon}
Return
Bit flip attack-guided mixed-precision neural network quantization
.png){kind=icon}
PRE
AI
DOI：10.1016/j.knosys.2025.114811.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Discover the advantageous impact of the bit flip attack on the mixed precision quantization. • Obtain the sensitivity indicator for each layer using the layer-by-layer bit flip attack. • Utilize the evolutionary algorithm to quickly search for the optimal mixed precision quantization strategy.