Matches in SemOpenAlex for { <https://semopenalex.org/work/W4224948081> ?p ?o ?g. }
Showing items 1 to 86 of
86
with 100 items per page.
- W4224948081 abstract "Language models generate texts by successively predicting probability distributions for next tokens given past ones. A growing field of interest tries to leverage external information in the decoding process so that the generated texts have desired properties, such as being more natural, non toxic, faithful, or having a specific writing style. A solution is to use a classifier at each generation step, resulting in a cooperative environment where the classifier guides the decoding of the language model distribution towards relevant texts for the task at hand. In this paper, we examine three families of (transformer-based) discriminators for this specific task of cooperative decoding: bidirectional, left-to-right and generative ones. We evaluate the pros and cons of these different types of discriminators for cooperative generation, exploring respective accuracy on classification tasks along with their impact on the resulting sample quality and computational performances. We also provide the code of a batched implementation of the powerful cooperative decoding strategy used for our experiments, the Monte Carlo Tree Search, working with each discriminator for Natural Language Generation." @default.
- W4224948081 created "2022-04-28" @default.
- W4224948081 creator A5009145141 @default.
- W4224948081 creator A5031635996 @default.
- W4224948081 creator A5036636249 @default.
- W4224948081 creator A5057169414 @default.
- W4224948081 creator A5082391569 @default.
- W4224948081 creator A5089419558 @default.
- W4224948081 creator A5090399761 @default.
- W4224948081 date "2022-07-06" @default.
- W4224948081 modified "2023-10-04" @default.
- W4224948081 title "Which Discriminator for Cooperative Text Generation?" @default.
- W4224948081 cites W2607151106 @default.
- W4224948081 cites W2766447205 @default.
- W4224948081 cites W2963283805 @default.
- W4224948081 cites W2963456134 @default.
- W4224948081 cites W2964268978 @default.
- W4224948081 cites W3096831136 @default.
- W4224948081 cites W3133702157 @default.
- W4224948081 cites W3155777744 @default.
- W4224948081 cites W3207707577 @default.
- W4224948081 cites W4205384019 @default.
- W4224948081 cites W4205393596 @default.
- W4224948081 doi "https://doi.org/10.1145/3477495.3531858" @default.
- W4224948081 hasPublicationYear "2022" @default.
- W4224948081 type Work @default.
- W4224948081 citedByCount "1" @default.
- W4224948081 crossrefType "proceedings-article" @default.
- W4224948081 hasAuthorship W4224948081A5009145141 @default.
- W4224948081 hasAuthorship W4224948081A5031635996 @default.
- W4224948081 hasAuthorship W4224948081A5036636249 @default.
- W4224948081 hasAuthorship W4224948081A5057169414 @default.
- W4224948081 hasAuthorship W4224948081A5082391569 @default.
- W4224948081 hasAuthorship W4224948081A5089419558 @default.
- W4224948081 hasAuthorship W4224948081A5090399761 @default.
- W4224948081 hasBestOaLocation W42249480812 @default.
- W4224948081 hasConcept C11413529 @default.
- W4224948081 hasConcept C119857082 @default.
- W4224948081 hasConcept C137293760 @default.
- W4224948081 hasConcept C153083717 @default.
- W4224948081 hasConcept C154945302 @default.
- W4224948081 hasConcept C195324797 @default.
- W4224948081 hasConcept C204321447 @default.
- W4224948081 hasConcept C2776187449 @default.
- W4224948081 hasConcept C2779803651 @default.
- W4224948081 hasConcept C39890363 @default.
- W4224948081 hasConcept C41008148 @default.
- W4224948081 hasConcept C57273362 @default.
- W4224948081 hasConcept C76155785 @default.
- W4224948081 hasConcept C94915269 @default.
- W4224948081 hasConcept C95623464 @default.
- W4224948081 hasConceptScore W4224948081C11413529 @default.
- W4224948081 hasConceptScore W4224948081C119857082 @default.
- W4224948081 hasConceptScore W4224948081C137293760 @default.
- W4224948081 hasConceptScore W4224948081C153083717 @default.
- W4224948081 hasConceptScore W4224948081C154945302 @default.
- W4224948081 hasConceptScore W4224948081C195324797 @default.
- W4224948081 hasConceptScore W4224948081C204321447 @default.
- W4224948081 hasConceptScore W4224948081C2776187449 @default.
- W4224948081 hasConceptScore W4224948081C2779803651 @default.
- W4224948081 hasConceptScore W4224948081C39890363 @default.
- W4224948081 hasConceptScore W4224948081C41008148 @default.
- W4224948081 hasConceptScore W4224948081C57273362 @default.
- W4224948081 hasConceptScore W4224948081C76155785 @default.
- W4224948081 hasConceptScore W4224948081C94915269 @default.
- W4224948081 hasConceptScore W4224948081C95623464 @default.
- W4224948081 hasLocation W42249480811 @default.
- W4224948081 hasLocation W42249480812 @default.
- W4224948081 hasLocation W42249480813 @default.
- W4224948081 hasLocation W42249480814 @default.
- W4224948081 hasLocation W42249480815 @default.
- W4224948081 hasOpenAccess W4224948081 @default.
- W4224948081 hasPrimaryLocation W42249480811 @default.
- W4224948081 hasRelatedWork W2412510955 @default.
- W4224948081 hasRelatedWork W2757136988 @default.
- W4224948081 hasRelatedWork W2783089240 @default.
- W4224948081 hasRelatedWork W2970526979 @default.
- W4224948081 hasRelatedWork W2971644172 @default.
- W4224948081 hasRelatedWork W3130493457 @default.
- W4224948081 hasRelatedWork W4221144361 @default.
- W4224948081 hasRelatedWork W4221144473 @default.
- W4224948081 hasRelatedWork W4280544492 @default.
- W4224948081 hasRelatedWork W4292402473 @default.
- W4224948081 isParatext "false" @default.
- W4224948081 isRetracted "false" @default.
- W4224948081 workType "article" @default.