Researchers detail a new type of attack that could increase the energy consumption of neural networks by adding small amounts of noise to a network's inputs (Karen Hao/MIT Technology Review)

Researchers detail a new type of attack that could increase the energy consumption of neural networks by adding small amounts of noise to a network's inputs (Karen Hao/MIT Technology Review)




Karen Hao / MIT Technology Review:

Researchers detail a new type of attack that could increase the energy consumption of neural networks by adding small amounts of noise to a network’s inputs  —  The latest generation of neural networks are vulnerable to a new kind of attack that makes them use too much energy.





Source link

Similar Articles

Comments

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Advertismentspot_img

Instagram

Most Popular