FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3091567777> ?p ?o ?g. }

• W3091567777 abstract "Contrastive approaches to representation learning have recently shown great promise. In contrast to generative approaches, these contrastive models learn a deterministic encoder with no notion of uncertainty or confidence. In this paper, we introduce a simple approach based on contrasting distributions that learns to assign uncertainty for pretrained contrastive representations. In particular, we train a deep network from a representation to a distribution in representation space, whose variance can be used as a measure of confidence. In our experiments, we show that this deep uncertainty model can be used (1) to visually interpret model behavior, (2) to detect new noise in the input to deployed models, (3) to detect anomalies, where we outperform 10 baseline methods across 11 tasks with improvements of up to 14% absolute, and (4) to classify out-of-distribution examples where our fully unsupervised model is competitive with supervised methods." @default.
• W3091567777 created "2020-10-08" @default.
• W3091567777 creator A5001961716 @default.
• W3091567777 creator A5089742311 @default.
• W3091567777 date "2021-05-04" @default.
• W3091567777 modified "2023-10-16" @default.
• W3091567777 title "A Simple Framework for Uncertainty in Contrastive Learning" @default.
• W3091567777 cites W1776456436 @default.
• W3091567777 cites W1787224781 @default.
• W3091567777 cites W1909320841 @default.
• W3091567777 cites W1959608418 @default.
• W3091567777 cites W1980262437 @default.
• W3091567777 cites W2041559880 @default.
• W3091567777 cites W2099471712 @default.
• W3091567777 cites W2101234009 @default.
• W3091567777 cites W2113060550 @default.
• W3091567777 cites W2117539524 @default.
• W3091567777 cites W2122538988 @default.
• W3091567777 cites W2136922672 @default.
• W3091567777 cites W2285233685 @default.
• W3091567777 cites W2296719434 @default.
• W3091567777 cites W2321533354 @default.
• W3091567777 cites W2326925005 @default.
• W3091567777 cites W2335728318 @default.
• W3091567777 cites W2403726878 @default.
• W3091567777 cites W2411541852 @default.
• W3091567777 cites W2412589713 @default.
• W3091567777 cites W241901432 @default.
• W3091567777 cites W2510089447 @default.
• W3091567777 cites W2531327146 @default.
• W3091567777 cites W2549418531 @default.
• W3091567777 cites W2575671312 @default.
• W3091567777 cites W2617934075 @default.
• W3091567777 cites W2622563070 @default.
• W3091567777 cites W2727650337 @default.
• W3091567777 cites W2785902771 @default.
• W3091567777 cites W2807807246 @default.
• W3091567777 cites W2807955733 @default.
• W3091567777 cites W2867167548 @default.
• W3091567777 cites W2904981516 @default.
• W3091567777 cites W2910068345 @default.
• W3091567777 cites W2911896682 @default.
• W3091567777 cites W2948140294 @default.
• W3091567777 cites W2949517790 @default.
• W3091567777 cites W2951873722 @default.
• W3091567777 cites W2962723986 @default.
• W3091567777 cites W2962742544 @default.
• W3091567777 cites W2962824366 @default.
• W3091567777 cites W2962858109 @default.
• W3091567777 cites W2963060032 @default.
• W3091567777 cites W2963198154 @default.
• W3091567777 cites W2963445059 @default.
• W3091567777 cites W2963466845 @default.
• W3091567777 cites W2963693742 @default.
• W3091567777 cites W2964121744 @default.
• W3091567777 cites W2971155163 @default.
• W3091567777 cites W2977162702 @default.
• W3091567777 cites W2989929945 @default.
• W3091567777 cites W2995253937 @default.
• W3091567777 cites W2998388430 @default.
• W3091567777 cites W3005680577 @default.
• W3091567777 cites W3009561768 @default.
• W3091567777 cites W3013557831 @default.
• W3091567777 cites W3029860052 @default.
• W3091567777 cites W3034781633 @default.
• W3091567777 cites W3035060554 @default.
• W3091567777 cites W3035524453 @default.
• W3091567777 cites W3106428938 @default.
• W3091567777 cites W3118608800 @default.
• W3091567777 cites W343636949 @default.
• W3091567777 cites W967544008 @default.
• W3091567777 hasPublicationYear "2021" @default.
• W3091567777 type Work @default.
• W3091567777 sameAs 3091567777 @default.
• W3091567777 citedByCount "2" @default.
• W3091567777 countsByYear W30915677772021 @default.
• W3091567777 crossrefType "journal-article" @default.
• W3091567777 hasAuthorship W3091567777A5001961716 @default.
• W3091567777 hasAuthorship W3091567777A5089742311 @default.
• W3091567777 hasConcept C111472728 @default.
• W3091567777 hasConcept C111919701 @default.
• W3091567777 hasConcept C115961682 @default.
• W3091567777 hasConcept C118505674 @default.
• W3091567777 hasConcept C119857082 @default.
• W3091567777 hasConcept C121955636 @default.
• W3091567777 hasConcept C138885662 @default.
• W3091567777 hasConcept C144133560 @default.
• W3091567777 hasConcept C153180895 @default.
• W3091567777 hasConcept C154945302 @default.
• W3091567777 hasConcept C167966045 @default.
• W3091567777 hasConcept C17744445 @default.
• W3091567777 hasConcept C196083921 @default.
• W3091567777 hasConcept C199539241 @default.
• W3091567777 hasConcept C204321447 @default.
• W3091567777 hasConcept C2776359362 @default.
• W3091567777 hasConcept C2776502983 @default.
• W3091567777 hasConcept C2780586882 @default.
• W3091567777 hasConcept C39890363 @default.
• W3091567777 hasConcept C41008148 @default.
• W3091567777 hasConcept C59404180 @default.

• next