General‎ > ‎

[News] Peering into neural networks

posted Jul 10, 2017, 10:02 PM by DaWoon Heo
  (Image: Christine Daniloff/MIT)
[ "New technique helps elucidate the inner workings of neural networks trained on visual data" ]

 Neural networks is for performing computational tasks by learning. The learning is conducted by analyze large sets of training data.
However, it is difficult to recognize which data they are processing between input and output.
A researchers from Computer Science and Artificial Intelligence Laboratory (CSAIL) of MIT, showed a method to find out the process during training the visual scenes identification. From this study, Bau, one of the researcher of this study, mentioned that it suggests that "neural networks are actually trying to approximate getting a grandmother neuron".

If you interested in, please click the following link.