FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3034642726> ?p ?o ?g. }

• W3034642726 abstract "While modern deep neural architectures generalise well when test data is sampled from the same distribution as training data, they fail badly for cases when the test data distribution differs from the training distribution even along a few dimensions. This lack of out-of-distribution generalisation is increasingly manifested when the tasks become more abstract and complex, such as in relational reasoning. In this paper we propose a neuroscience-inspired inductive-biased module that can be readily amalgamated with current neural network architectures to improve out-of-distribution (o.o.d) generalisation performance on relational reasoning tasks. This module learns to project high-dimensional object representations to low-dimensional manifolds for more efficient and generalisable relational comparisons. We show that neural nets with this inductive bias achieve considerably better o.o.d generalisation performance for a range of relational reasoning tasks. We finally analyse the proposed inductive bias module to understand the importance of lower dimension projection, and propose an augmentation to the algorithmic alignment theory to better measure algorithmic alignment with generalisation." @default.
• W3034642726 created "2020-06-19" @default.
• W3034642726 creator A5011740583 @default.
• W3034642726 creator A5036018012 @default.
• W3034642726 creator A5056748708 @default.
• W3034642726 date "2020-06-15" @default.
• W3034642726 modified "2023-09-24" @default.
• W3034642726 title "Extrapolatable Relational Reasoning With Comparators in Low-Dimensional Manifolds" @default.
• W3034642726 cites W1538131130 @default.
• W3034642726 cites W1715013381 @default.
• W3034642726 cites W1933349210 @default.
• W3034642726 cites W1970792572 @default.
• W3034642726 cites W2076432979 @default.
• W3034642726 cites W2089468765 @default.
• W3034642726 cites W2090928435 @default.
• W3034642726 cites W2099471712 @default.
• W3034642726 cites W2099741732 @default.
• W3034642726 cites W2103212315 @default.
• W3034642726 cites W2510108431 @default.
• W3034642726 cites W2561715562 @default.
• W3034642726 cites W2617995259 @default.
• W3034642726 cites W2753738274 @default.
• W3034642726 cites W2763549966 @default.
• W3034642726 cites W2765345568 @default.
• W3034642726 cites W2787337315 @default.
• W3034642726 cites W2804072623 @default.
• W3034642726 cites W2804180102 @default.
• W3034642726 cites W2805516822 @default.
• W3034642726 cites W2887020936 @default.
• W3034642726 cites W2888339491 @default.
• W3034642726 cites W2891021031 @default.
• W3034642726 cites W2899530572 @default.
• W3034642726 cites W2919112510 @default.
• W3034642726 cites W2945286741 @default.
• W3034642726 cites W2949737767 @default.
• W3034642726 cites W2950761309 @default.
• W3034642726 cites W2951004968 @default.
• W3034642726 cites W2952433032 @default.
• W3034642726 cites W2953273646 @default.
• W3034642726 cites W2953388933 @default.
• W3034642726 cites W2953494151 @default.
• W3034642726 cites W2955554316 @default.
• W3034642726 cites W2958360136 @default.
• W3034642726 cites W2962900737 @default.
• W3034642726 cites W2963403868 @default.
• W3034642726 cites W2963486392 @default.
• W3034642726 cites W2963661177 @default.
• W3034642726 cites W2963907629 @default.
• W3034642726 cites W2968917279 @default.
• W3034642726 cites W2971044234 @default.
• W3034642726 cites W2971127577 @default.
• W3034642726 cites W2971200848 @default.
• W3034642726 cites W2981514504 @default.
• W3034642726 cites W2989818084 @default.
• W3034642726 cites W2992535490 @default.
• W3034642726 cites W2994958563 @default.
• W3034642726 cites W2995679616 @default.
• W3034642726 cites W2995889583 @default.
• W3034642726 cites W2999956363 @default.
• W3034642726 cites W3007978994 @default.
• W3034642726 cites W3014016778 @default.
• W3034642726 cites W3035349046 @default.
• W3034642726 cites W3108961219 @default.
• W3034642726 doi "https://doi.org/10.48550/arxiv.2006.08698" @default.
• W3034642726 hasPublicationYear "2020" @default.
• W3034642726 type Work @default.
• W3034642726 sameAs 3034642726 @default.
• W3034642726 citedByCount "1" @default.
• W3034642726 countsByYear W30346427262020 @default.
• W3034642726 crossrefType "posted-content" @default.
• W3034642726 hasAuthorship W3034642726A5011740583 @default.
• W3034642726 hasAuthorship W3034642726A5036018012 @default.
• W3034642726 hasAuthorship W3034642726A5056748708 @default.
• W3034642726 hasBestOaLocation W30346427261 @default.
• W3034642726 hasConcept C110121322 @default.
• W3034642726 hasConcept C11413529 @default.
• W3034642726 hasConcept C119857082 @default.
• W3034642726 hasConcept C134306372 @default.
• W3034642726 hasConcept C154945302 @default.
• W3034642726 hasConcept C159985019 @default.
• W3034642726 hasConcept C162324750 @default.
• W3034642726 hasConcept C187736073 @default.
• W3034642726 hasConcept C192562407 @default.
• W3034642726 hasConcept C197352929 @default.
• W3034642726 hasConcept C202444582 @default.
• W3034642726 hasConcept C204323151 @default.
• W3034642726 hasConcept C2780451532 @default.
• W3034642726 hasConcept C2781238097 @default.
• W3034642726 hasConcept C28006648 @default.
• W3034642726 hasConcept C2984842247 @default.
• W3034642726 hasConcept C33676613 @default.
• W3034642726 hasConcept C33923547 @default.
• W3034642726 hasConcept C41008148 @default.
• W3034642726 hasConcept C50644808 @default.
• W3034642726 hasConcept C57493831 @default.
• W3034642726 hasConcept C80444323 @default.
• W3034642726 hasConceptScore W3034642726C110121322 @default.
• W3034642726 hasConceptScore W3034642726C11413529 @default.
• W3034642726 hasConceptScore W3034642726C119857082 @default.
• W3034642726 hasConceptScore W3034642726C134306372 @default.

• next