Matches in SemOpenAlex for { <https://semopenalex.org/work/W3042177413> ?p ?o ?g. }
- W3042177413 abstract "The observation of sparse trainable sub-networks within over-parametrized networks - also known as Lottery Tickets (LTs) - has prompted inquiries around their trainability, scaling, uniqueness, and generalization properties. Across 28 combinations of image classification tasks and architectures, we discover differences in the connectivity structure of LTs found through different iterative pruning techniques, thus disproving their uniqueness and connecting emergent mask structure to the choice of pruning. In addition, we propose a consensus-based method for generating refined lottery tickets. This lottery ticket denoising procedure, based on the principle that parameters that always go unpruned across different tasks more reliably identify important sub-networks, is capable of selecting a meaningful portion of the architecture in an embarrassingly parallel way, while quickly discarding extra parameters without the need for further pruning iterations. We successfully train these sub-networks to performance comparable to that of ordinary lottery tickets." @default.
- W3042177413 created "2020-07-16" @default.
- W3042177413 creator A5059618221 @default.
- W3042177413 creator A5089018232 @default.
- W3042177413 date "2020-07-06" @default.
- W3042177413 modified "2023-10-03" @default.
- W3042177413 title "Bespoke vs. Prêt-à-Porter Lottery Tickets: Exploiting Mask Similarity for Trainable Sub-Network Finding." @default.
- W3042177413 cites W156878426 @default.
- W3042177413 cites W1604329830 @default.
- W3042177413 cites W1677182931 @default.
- W3042177413 cites W1686810756 @default.
- W3042177413 cites W2111935653 @default.
- W3042177413 cites W2114766824 @default.
- W3042177413 cites W2117539524 @default.
- W3042177413 cites W2119144962 @default.
- W3042177413 cites W2120972216 @default.
- W3042177413 cites W2125389748 @default.
- W3042177413 cites W2154579312 @default.
- W3042177413 cites W2163605009 @default.
- W3042177413 cites W2171658832 @default.
- W3042177413 cites W2335728318 @default.
- W3042177413 cites W2554302513 @default.
- W3042177413 cites W2590796488 @default.
- W3042177413 cites W2605409611 @default.
- W3042177413 cites W2742056655 @default.
- W3042177413 cites W2750384547 @default.
- W3042177413 cites W2796664995 @default.
- W3042177413 cites W2806956262 @default.
- W3042177413 cites W2894604724 @default.
- W3042177413 cites W2899771611 @default.
- W3042177413 cites W2902986194 @default.
- W3042177413 cites W2907127169 @default.
- W3042177413 cites W2915589364 @default.
- W3042177413 cites W2919347051 @default.
- W3042177413 cites W2923764619 @default.
- W3042177413 cites W2943283198 @default.
- W3042177413 cites W2943925420 @default.
- W3042177413 cites W2948130861 @default.
- W3042177413 cites W2948253213 @default.
- W3042177413 cites W2948635472 @default.
- W3042177413 cites W2949117887 @default.
- W3042177413 cites W2949650786 @default.
- W3042177413 cites W2951569836 @default.
- W3042177413 cites W2962662229 @default.
- W3042177413 cites W2962930448 @default.
- W3042177413 cites W2963064482 @default.
- W3042177413 cites W2963319203 @default.
- W3042177413 cites W2963403868 @default.
- W3042177413 cites W2963674932 @default.
- W3042177413 cites W2963759780 @default.
- W3042177413 cites W2963813662 @default.
- W3042177413 cites W2963981420 @default.
- W3042177413 cites W2964233199 @default.
- W3042177413 cites W2970330753 @default.
- W3042177413 cites W2972810859 @default.
- W3042177413 cites W2981985696 @default.
- W3042177413 cites W2996594819 @default.
- W3042177413 cites W3023169080 @default.
- W3042177413 cites W3023648375 @default.
- W3042177413 cites W3030163527 @default.
- W3042177413 cites W3118608800 @default.
- W3042177413 cites W3167660944 @default.
- W3042177413 hasPublicationYear "2020" @default.
- W3042177413 type Work @default.
- W3042177413 sameAs 3042177413 @default.
- W3042177413 citedByCount "1" @default.
- W3042177413 countsByYear W30421774132021 @default.
- W3042177413 crossrefType "posted-content" @default.
- W3042177413 hasAuthorship W3042177413A5059618221 @default.
- W3042177413 hasAuthorship W3042177413A5089018232 @default.
- W3042177413 hasConcept C103278499 @default.
- W3042177413 hasConcept C105795698 @default.
- W3042177413 hasConcept C108010975 @default.
- W3042177413 hasConcept C111335779 @default.
- W3042177413 hasConcept C115961682 @default.
- W3042177413 hasConcept C119857082 @default.
- W3042177413 hasConcept C134306372 @default.
- W3042177413 hasConcept C154945302 @default.
- W3042177413 hasConcept C177148314 @default.
- W3042177413 hasConcept C2524010 @default.
- W3042177413 hasConcept C2777021972 @default.
- W3042177413 hasConcept C2777340749 @default.
- W3042177413 hasConcept C33923547 @default.
- W3042177413 hasConcept C41008148 @default.
- W3042177413 hasConcept C6557445 @default.
- W3042177413 hasConcept C80444323 @default.
- W3042177413 hasConcept C86803240 @default.
- W3042177413 hasConceptScore W3042177413C103278499 @default.
- W3042177413 hasConceptScore W3042177413C105795698 @default.
- W3042177413 hasConceptScore W3042177413C108010975 @default.
- W3042177413 hasConceptScore W3042177413C111335779 @default.
- W3042177413 hasConceptScore W3042177413C115961682 @default.
- W3042177413 hasConceptScore W3042177413C119857082 @default.
- W3042177413 hasConceptScore W3042177413C134306372 @default.
- W3042177413 hasConceptScore W3042177413C154945302 @default.
- W3042177413 hasConceptScore W3042177413C177148314 @default.
- W3042177413 hasConceptScore W3042177413C2524010 @default.
- W3042177413 hasConceptScore W3042177413C2777021972 @default.
- W3042177413 hasConceptScore W3042177413C2777340749 @default.
- W3042177413 hasConceptScore W3042177413C33923547 @default.