FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3165377449> ?p ?o ?g. }

• W3165377449 abstract "Masking some input variables of a deep neural network (DNN) and computing output changes on the masked input sample represent a typical way to compute attributions of input variables in the sample. People usually mask an input variable using its baseline value. However, there is no theory to examine whether baseline value faithfully represents the absence of an input variable, emph{i.e.,} removing all signals from the input variable. Fortunately, recent studies show that the inference score of a DNN can be strictly disentangled into a set of causal patterns (or concepts) encoded by the DNN. Therefore, we propose to use causal patterns to examine the faithfulness of baseline values. More crucially, it is proven that causal patterns can be explained as the elementary rationale of the Shapley value. Furthermore, we propose a method to learn optimal baseline values, and experimental results have demonstrated its effectiveness." @default.
• W3165377449 created "2021-06-07" @default.
• W3165377449 creator A5011795597 @default.
• W3165377449 creator A5027523191 @default.
• W3165377449 creator A5051387840 @default.
• W3165377449 creator A5080312761 @default.
• W3165377449 date "2021-05-22" @default.
• W3165377449 modified "2023-09-25" @default.
• W3165377449 title "Can We Faithfully Represent Masked States to Compute Shapley Values on a DNN?" @default.
• W3165377449 cites W1562353621 @default.
• W3165377449 cites W1825675169 @default.
• W3165377449 cites W1967638691 @default.
• W3165377449 cites W1995334767 @default.
• W3165377449 cites W1998095641 @default.
• W3165377449 cites W2009662043 @default.
• W3165377449 cites W2055202133 @default.
• W3165377449 cites W2073231946 @default.
• W3165377449 cites W2081822233 @default.
• W3165377449 cites W2097310377 @default.
• W3165377449 cites W2112796928 @default.
• W3165377449 cites W2129888542 @default.
• W3165377449 cites W2155279768 @default.
• W3165377449 cites W2163605009 @default.
• W3165377449 cites W2169393322 @default.
• W3165377449 cites W2194775991 @default.
• W3165377449 cites W2295107390 @default.
• W3165377449 cites W2594633041 @default.
• W3165377449 cites W2619351609 @default.
• W3165377449 cites W2787070805 @default.
• W3165377449 cites W2803532212 @default.
• W3165377449 cites W2898422183 @default.
• W3165377449 cites W2923915248 @default.
• W3165377449 cites W2924181074 @default.
• W3165377449 cites W2962858109 @default.
• W3165377449 cites W2962862931 @default.
• W3165377449 cites W2963029978 @default.
• W3165377449 cites W2963424533 @default.
• W3165377449 cites W2963989815 @default.
• W3165377449 cites W2963996492 @default.
• W3165377449 cites W2964159526 @default.
• W3165377449 cites W2964253222 @default.
• W3165377449 cites W2988157455 @default.
• W3165377449 cites W2989972208 @default.
• W3165377449 cites W2996507500 @default.
• W3165377449 cites W3005086430 @default.
• W3165377449 cites W3005535506 @default.
• W3165377449 cites W3007549203 @default.
• W3165377449 cites W3034522874 @default.
• W3165377449 cites W3034764457 @default.
• W3165377449 cites W3035235410 @default.
• W3165377449 cites W3092088514 @default.
• W3165377449 cites W3098134138 @default.
• W3165377449 cites W3101609372 @default.
• W3165377449 cites W3106080991 @default.
• W3165377449 cites W3118608800 @default.
• W3165377449 cites W3121397038 @default.
• W3165377449 cites W3124464611 @default.
• W3165377449 cites W3138813434 @default.
• W3165377449 cites W3198754983 @default.
• W3165377449 doi "https://doi.org/10.48550/arxiv.2105.10719" @default.
• W3165377449 hasPublicationYear "2021" @default.
• W3165377449 type Work @default.
• W3165377449 sameAs 3165377449 @default.
• W3165377449 citedByCount "4" @default.
• W3165377449 countsByYear W31653774492021 @default.
• W3165377449 crossrefType "posted-content" @default.
• W3165377449 hasAuthorship W3165377449A5011795597 @default.
• W3165377449 hasAuthorship W3165377449A5027523191 @default.
• W3165377449 hasAuthorship W3165377449A5051387840 @default.
• W3165377449 hasAuthorship W3165377449A5080312761 @default.
• W3165377449 hasBestOaLocation W31653774491 @default.
• W3165377449 hasConcept C105795698 @default.
• W3165377449 hasConcept C111368507 @default.
• W3165377449 hasConcept C11413529 @default.
• W3165377449 hasConcept C11671645 @default.
• W3165377449 hasConcept C119857082 @default.
• W3165377449 hasConcept C121332964 @default.
• W3165377449 hasConcept C12725497 @default.
• W3165377449 hasConcept C127313418 @default.
• W3165377449 hasConcept C134306372 @default.
• W3165377449 hasConcept C137002209 @default.
• W3165377449 hasConcept C142362112 @default.
• W3165377449 hasConcept C153349607 @default.
• W3165377449 hasConcept C154945302 @default.
• W3165377449 hasConcept C158600405 @default.
• W3165377449 hasConcept C177264268 @default.
• W3165377449 hasConcept C182365436 @default.
• W3165377449 hasConcept C185592680 @default.
• W3165377449 hasConcept C198531522 @default.
• W3165377449 hasConcept C199360897 @default.
• W3165377449 hasConcept C2776214188 @default.
• W3165377449 hasConcept C2776291640 @default.
• W3165377449 hasConcept C2777402240 @default.
• W3165377449 hasConcept C33923547 @default.
• W3165377449 hasConcept C41008148 @default.
• W3165377449 hasConcept C43617362 @default.
• W3165377449 hasConcept C50644808 @default.
• W3165377449 hasConcept C62520636 @default.
• W3165377449 hasConcept C84114770 @default.
• W3165377449 hasConceptScore W3165377449C105795698 @default.

• next