|
|
|
|
LEADER |
01000naa a22002652 4500 |
001 |
NLM292960689 |
003 |
DE-627 |
005 |
20231225074253.0 |
007 |
cr uuu---uuuuu |
008 |
231225s2018 xx |||||o 00| ||eng c |
024 |
7 |
|
|a 10.1109/MCG.2018.2878902
|2 doi
|
028 |
5 |
2 |
|a pubmed24n0976.xml
|
035 |
|
|
|a (DE-627)NLM292960689
|
035 |
|
|
|a (NLM)30668454
|
040 |
|
|
|a DE-627
|b ger
|c DE-627
|e rakwb
|
041 |
|
|
|a eng
|
100 |
1 |
|
|a Cashman, Dylan
|e verfasserin
|4 aut
|
245 |
1 |
0 |
|a RNNbow
|b Visualizing Learning Via Backpropagation Gradients in RNNs
|
264 |
|
1 |
|c 2018
|
336 |
|
|
|a Text
|b txt
|2 rdacontent
|
337 |
|
|
|a ƒaComputermedien
|b c
|2 rdamedia
|
338 |
|
|
|a ƒa Online-Ressource
|b cr
|2 rdacarrier
|
500 |
|
|
|a Date Completed 19.03.2019
|
500 |
|
|
|a Date Revised 19.03.2019
|
500 |
|
|
|a published: Print
|
500 |
|
|
|a Citation Status PubMed-not-MEDLINE
|
520 |
|
|
|a We present RNNbow, an interactive tool for visualizing the gradient flow during backpropagation in training of recurrent neural networks. By visualizing the gradient, as opposed to activations, RNNbow offers insight into how the network is learning. We show how it illustrates the vanishing gradient and the training process
|
650 |
|
4 |
|a Journal Article
|
650 |
|
4 |
|a Research Support, U.S. Gov't, Non-P.H.S.
|
700 |
1 |
|
|a Patterson, Genevieve
|e verfasserin
|4 aut
|
700 |
1 |
|
|a Mosca, Abigail
|e verfasserin
|4 aut
|
700 |
1 |
|
|a Watts, Nathan
|e verfasserin
|4 aut
|
700 |
1 |
|
|a Robinson, Shannon
|e verfasserin
|4 aut
|
700 |
1 |
|
|a Chang, Remco
|e verfasserin
|4 aut
|
773 |
0 |
8 |
|i Enthalten in
|t IEEE computer graphics and applications
|d 1991
|g 38(2018), 6 vom: 01. Nov., Seite 39-50
|w (DE-627)NLM098172794
|x 1558-1756
|7 nnns
|
773 |
1 |
8 |
|g volume:38
|g year:2018
|g number:6
|g day:01
|g month:11
|g pages:39-50
|
856 |
4 |
0 |
|u http://dx.doi.org/10.1109/MCG.2018.2878902
|3 Volltext
|
912 |
|
|
|a GBV_USEFLAG_A
|
912 |
|
|
|a SYSFLAG_A
|
912 |
|
|
|a GBV_NLM
|
912 |
|
|
|a GBV_ILN_350
|
951 |
|
|
|a AR
|
952 |
|
|
|d 38
|j 2018
|e 6
|b 01
|c 11
|h 39-50
|