TY - CONF
T1 - A JOHNSON-LINDENSTRAUSS FRAMEWORK FOR RANDOMLY INITIALIZED CNNS
AU - Nachum, Ido
AU - Hazła, Jan
AU - Gastpar, Michael
AU - Khina, Anatoly
N1 - Publisher Copyright:
© 2022 ICLR 2022 - 10th International Conference on Learning Representationss. All rights reserved.
PY - 2022
Y1 - 2022
N2 - How does the geometric representation of a dataset change after the application of each randomly initialized layer of a neural network? The celebrated Johnson-Lindenstrauss lemma answers this question for linear fully-connected neural networks (FNNs), stating that the geometry is essentially preserved. For FNNs with the ReLU activation, the angle between two inputs contracts according to a known mapping. The question for non-linear convolutional neural networks (CNNs) becomes much more intricate. To answer this question, we introduce a geometric framework. For linear CNNs, we show that the Johnson-Lindenstrauss lemma continues to hold, namely, that the angle between two inputs is preserved. For CNNs with ReLU activation, on the other hand, the behavior is richer: The angle between the outputs contracts, where the level of contraction depends on the nature of the inputs. In particular, after one layer, the geometry of natural images is essentially preserved, whereas for Gaussian correlated inputs, CNNs exhibit the same contracting behavior as FNNs with ReLU activation.
AB - How does the geometric representation of a dataset change after the application of each randomly initialized layer of a neural network? The celebrated Johnson-Lindenstrauss lemma answers this question for linear fully-connected neural networks (FNNs), stating that the geometry is essentially preserved. For FNNs with the ReLU activation, the angle between two inputs contracts according to a known mapping. The question for non-linear convolutional neural networks (CNNs) becomes much more intricate. To answer this question, we introduce a geometric framework. For linear CNNs, we show that the Johnson-Lindenstrauss lemma continues to hold, namely, that the angle between two inputs is preserved. For CNNs with ReLU activation, on the other hand, the behavior is richer: The angle between the outputs contracts, where the level of contraction depends on the nature of the inputs. In particular, after one layer, the geometry of natural images is essentially preserved, whereas for Gaussian correlated inputs, CNNs exhibit the same contracting behavior as FNNs with ReLU activation.
UR - http://www.scopus.com/inward/record.url?scp=85150369596&partnerID=8YFLogxK
M3 - ???researchoutput.researchoutputtypes.contributiontoconference.paper???
AN - SCOPUS:85150369596
T2 - 10th International Conference on Learning Representations, ICLR 2022
Y2 - 25 April 2022 through 29 April 2022
ER -