Matches in SemOpenAlex for { <https://semopenalex.org/work/W3108675487> ?p ?o ?g. }
Showing items 1 to 81 of
81
with 100 items per page.
- W3108675487 abstract "The recent success of machine learning systems on various QA datasets could be interpreted as a significant improvement in models' language understanding abilities. However, using various perturbations, multiple recent works have shown that good performance on a dataset might not indicate performance that correlates well with human's expectations from models that understand language. In this work we consider a top performing model on several Multiple Choice Question Answering (MCQA) datasets, and evaluate it against a set of expectations one might have from such a model, using a series of zero-information perturbations of the model's inputs. Our results show that the model clearly falls short of our expectations, and motivates a modified training approach that forces the model to better attend to the inputs. We show that the new training paradigm leads to a model that performs on par with the original model while better satisfying our expectations." @default.
- W3108675487 created "2020-12-07" @default.
- W3108675487 creator A5023802054 @default.
- W3108675487 creator A5053062856 @default.
- W3108675487 creator A5082938551 @default.
- W3108675487 date "2020-11-20" @default.
- W3108675487 modified "2023-09-23" @default.
- W3108675487 title "What do we expect from Multiple-choice QA Systems?" @default.
- W3108675487 cites W2337282450 @default.
- W3108675487 cites W2606964149 @default.
- W3108675487 cites W2794325560 @default.
- W3108675487 cites W2890894339 @default.
- W3108675487 cites W2962727366 @default.
- W3108675487 cites W2962736243 @default.
- W3108675487 cites W2963545917 @default.
- W3108675487 cites W2963969878 @default.
- W3108675487 cites W2964048171 @default.
- W3108675487 cites W2965373594 @default.
- W3108675487 cites W2970476646 @default.
- W3108675487 cites W2987215000 @default.
- W3108675487 cites W2994934025 @default.
- W3108675487 cites W2996848635 @default.
- W3108675487 cites W2996851481 @default.
- W3108675487 cites W3015339932 @default.
- W3108675487 cites W3035331128 @default.
- W3108675487 cites W3105261549 @default.
- W3108675487 doi "https://doi.org/10.48550/arxiv.2011.10647" @default.
- W3108675487 hasPublicationYear "2020" @default.
- W3108675487 type Work @default.
- W3108675487 sameAs 3108675487 @default.
- W3108675487 citedByCount "0" @default.
- W3108675487 crossrefType "posted-content" @default.
- W3108675487 hasAuthorship W3108675487A5023802054 @default.
- W3108675487 hasAuthorship W3108675487A5053062856 @default.
- W3108675487 hasAuthorship W3108675487A5082938551 @default.
- W3108675487 hasBestOaLocation W31086754871 @default.
- W3108675487 hasConcept C119857082 @default.
- W3108675487 hasConcept C137293760 @default.
- W3108675487 hasConcept C138885662 @default.
- W3108675487 hasConcept C143724316 @default.
- W3108675487 hasConcept C151730666 @default.
- W3108675487 hasConcept C154945302 @default.
- W3108675487 hasConcept C177264268 @default.
- W3108675487 hasConcept C199360897 @default.
- W3108675487 hasConcept C2780813799 @default.
- W3108675487 hasConcept C41008148 @default.
- W3108675487 hasConcept C41895202 @default.
- W3108675487 hasConcept C44291984 @default.
- W3108675487 hasConcept C51632099 @default.
- W3108675487 hasConcept C86803240 @default.
- W3108675487 hasConceptScore W3108675487C119857082 @default.
- W3108675487 hasConceptScore W3108675487C137293760 @default.
- W3108675487 hasConceptScore W3108675487C138885662 @default.
- W3108675487 hasConceptScore W3108675487C143724316 @default.
- W3108675487 hasConceptScore W3108675487C151730666 @default.
- W3108675487 hasConceptScore W3108675487C154945302 @default.
- W3108675487 hasConceptScore W3108675487C177264268 @default.
- W3108675487 hasConceptScore W3108675487C199360897 @default.
- W3108675487 hasConceptScore W3108675487C2780813799 @default.
- W3108675487 hasConceptScore W3108675487C41008148 @default.
- W3108675487 hasConceptScore W3108675487C41895202 @default.
- W3108675487 hasConceptScore W3108675487C44291984 @default.
- W3108675487 hasConceptScore W3108675487C51632099 @default.
- W3108675487 hasConceptScore W3108675487C86803240 @default.
- W3108675487 hasLocation W31086754871 @default.
- W3108675487 hasOpenAccess W3108675487 @default.
- W3108675487 hasPrimaryLocation W31086754871 @default.
- W3108675487 hasRelatedWork W10719664 @default.
- W3108675487 hasRelatedWork W11356396 @default.
- W3108675487 hasRelatedWork W11464338 @default.
- W3108675487 hasRelatedWork W12824513 @default.
- W3108675487 hasRelatedWork W13451536 @default.
- W3108675487 hasRelatedWork W13607926 @default.
- W3108675487 hasRelatedWork W193554 @default.
- W3108675487 hasRelatedWork W317670 @default.
- W3108675487 hasRelatedWork W5769024 @default.
- W3108675487 hasRelatedWork W7982726 @default.
- W3108675487 isParatext "false" @default.
- W3108675487 isRetracted "false" @default.
- W3108675487 magId "3108675487" @default.
- W3108675487 workType "article" @default.