Matches in SemOpenAlex for { <https://semopenalex.org/work/W2395582425> ?p ?o ?g. }
Showing items 1 to 87 of
87
with 100 items per page.
- W2395582425 abstract "Feature representations in both, biological neural networks in the primate ventral stream and artificial convolutional neural networks trained on object recognition, incresase in complexity and receptive field size with layer depth. Somewhat strikingly, empirical evidence indicates that this analogy extends to the specific representations learned in each layer. This suggests that biological and artificial neural networks share a fundamental organising principle. We shed light on this principle in the framework of optimal coding. Specifically, we first investigate which properties of a code render it robust to transmission over noisy channels and formally prove that for equientropic channels an upper bound on the expected minimum decoding error is attained for codes with maximum marginal entropy. We then show that the pairwise correlation of units in a deep layer of a neural network, that has been trained on an object recognition task, increases when perturbing the distribution of input images, i. e., that the network exhibits properties of an optimally coding system. By analogy, this suggests that the layer-wise similarity of feature representations in biological and artificial neural networks is a result of optimal coding that enables robust transmission of object information over noisy channels. Because we find that in equientropic channels the upper bound on the expected minimum decoding error is independent of the class-conditional entropy, our work further provides a plausible explanation why optimal codes can be learned in unsupervised settings." @default.
- W2395582425 created "2016-06-24" @default.
- W2395582425 creator A5040565095 @default.
- W2395582425 creator A5044005697 @default.
- W2395582425 creator A5064364907 @default.
- W2395582425 creator A5075255072 @default.
- W2395582425 date "2016-05-23" @default.
- W2395582425 modified "2023-09-27" @default.
- W2395582425 title "Optimal Coding in Biological and Artificial Neural Networks." @default.
- W2395582425 cites W144689751 @default.
- W2395582425 cites W1995875735 @default.
- W2395582425 cites W2099294566 @default.
- W2395582425 cites W2107366978 @default.
- W2395582425 cites W2117539524 @default.
- W2395582425 cites W2121326391 @default.
- W2395582425 cites W2166206801 @default.
- W2395582425 cites W81409850 @default.
- W2395582425 hasPublicationYear "2016" @default.
- W2395582425 type Work @default.
- W2395582425 sameAs 2395582425 @default.
- W2395582425 citedByCount "0" @default.
- W2395582425 crossrefType "posted-content" @default.
- W2395582425 hasAuthorship W2395582425A5040565095 @default.
- W2395582425 hasAuthorship W2395582425A5044005697 @default.
- W2395582425 hasAuthorship W2395582425A5064364907 @default.
- W2395582425 hasAuthorship W2395582425A5075255072 @default.
- W2395582425 hasConcept C105795698 @default.
- W2395582425 hasConcept C106301342 @default.
- W2395582425 hasConcept C11413529 @default.
- W2395582425 hasConcept C121332964 @default.
- W2395582425 hasConcept C134306372 @default.
- W2395582425 hasConcept C153180895 @default.
- W2395582425 hasConcept C154945302 @default.
- W2395582425 hasConcept C179518139 @default.
- W2395582425 hasConcept C184898388 @default.
- W2395582425 hasConcept C33923547 @default.
- W2395582425 hasConcept C41008148 @default.
- W2395582425 hasConcept C50644808 @default.
- W2395582425 hasConcept C57273362 @default.
- W2395582425 hasConcept C62520636 @default.
- W2395582425 hasConcept C77553402 @default.
- W2395582425 hasConcept C80444323 @default.
- W2395582425 hasConcept C81363708 @default.
- W2395582425 hasConceptScore W2395582425C105795698 @default.
- W2395582425 hasConceptScore W2395582425C106301342 @default.
- W2395582425 hasConceptScore W2395582425C11413529 @default.
- W2395582425 hasConceptScore W2395582425C121332964 @default.
- W2395582425 hasConceptScore W2395582425C134306372 @default.
- W2395582425 hasConceptScore W2395582425C153180895 @default.
- W2395582425 hasConceptScore W2395582425C154945302 @default.
- W2395582425 hasConceptScore W2395582425C179518139 @default.
- W2395582425 hasConceptScore W2395582425C184898388 @default.
- W2395582425 hasConceptScore W2395582425C33923547 @default.
- W2395582425 hasConceptScore W2395582425C41008148 @default.
- W2395582425 hasConceptScore W2395582425C50644808 @default.
- W2395582425 hasConceptScore W2395582425C57273362 @default.
- W2395582425 hasConceptScore W2395582425C62520636 @default.
- W2395582425 hasConceptScore W2395582425C77553402 @default.
- W2395582425 hasConceptScore W2395582425C80444323 @default.
- W2395582425 hasConceptScore W2395582425C81363708 @default.
- W2395582425 hasLocation W23955824251 @default.
- W2395582425 hasOpenAccess W2395582425 @default.
- W2395582425 hasPrimaryLocation W23955824251 @default.
- W2395582425 hasRelatedWork W1550736670 @default.
- W2395582425 hasRelatedWork W1700942727 @default.
- W2395582425 hasRelatedWork W1975512047 @default.
- W2395582425 hasRelatedWork W2056501338 @default.
- W2395582425 hasRelatedWork W2161388792 @default.
- W2395582425 hasRelatedWork W2305996987 @default.
- W2395582425 hasRelatedWork W2385174988 @default.
- W2395582425 hasRelatedWork W2540404261 @default.
- W2395582425 hasRelatedWork W2618540739 @default.
- W2395582425 hasRelatedWork W2805149973 @default.
- W2395582425 hasRelatedWork W2898737887 @default.
- W2395582425 hasRelatedWork W2964324450 @default.
- W2395582425 hasRelatedWork W2981710180 @default.
- W2395582425 hasRelatedWork W3003652842 @default.
- W2395582425 hasRelatedWork W3003893721 @default.
- W2395582425 hasRelatedWork W3049639017 @default.
- W2395582425 hasRelatedWork W3102048404 @default.
- W2395582425 hasRelatedWork W3136498298 @default.
- W2395582425 hasRelatedWork W3197904945 @default.
- W2395582425 hasRelatedWork W3200580033 @default.
- W2395582425 isParatext "false" @default.
- W2395582425 isRetracted "false" @default.
- W2395582425 magId "2395582425" @default.
- W2395582425 workType "article" @default.