Matches in SemOpenAlex for { <https://semopenalex.org/work/W4385570801> ?p ?o ?g. }
Showing items 1 to 77 of
77
with 100 items per page.
- W4385570801 abstract "Pre-trained large language models (PLMs) underly most new developments in natural language processing. They have shifted the field from application-specific model pipelines to a single model that is adapted to a wide range of tasks. Autoregressive PLMs like GPT-3 or PaLM and associated techniques like fewshot learning, have additionally shifted the output modality to generation instead of classification or regression. Despite their ubiquitous use, the generation quality of language models is rarely evaluated when these models are introduced. Additionally, it is unclear how existing generation tasks–while they can be used to compare systems at a high level–relate to the real world use cases for which people have been adopting them. In this work, we discuss how to adapt existing application-specific generation benchmarks to PLMs and provide an in-depth, empirical study of the limitations and capabilities of PLMs in natural language generation tasks along dimensions such as scale, architecture, input and output language. Our results show that PLMs differ in their applicability to different data regimes and their generalization to multiple languages. They further inform practitioners as to which PLMs to use for a given generation task setup. We share best practices to be taken into consideration when benchmarking generation capabilities during the development of upcoming PLMs." @default.
- W4385570801 created "2023-08-05" @default.
- W4385570801 creator A5021334672 @default.
- W4385570801 creator A5026789962 @default.
- W4385570801 creator A5047179830 @default.
- W4385570801 date "2023-01-01" @default.
- W4385570801 modified "2023-09-24" @default.
- W4385570801 title "Benchmarking Large Language Model Capabilities for Conditional Generation" @default.
- W4385570801 doi "https://doi.org/10.18653/v1/2023.acl-long.511" @default.
- W4385570801 hasPublicationYear "2023" @default.
- W4385570801 type Work @default.
- W4385570801 citedByCount "0" @default.
- W4385570801 crossrefType "proceedings-article" @default.
- W4385570801 hasAuthorship W4385570801A5021334672 @default.
- W4385570801 hasAuthorship W4385570801A5026789962 @default.
- W4385570801 hasAuthorship W4385570801A5047179830 @default.
- W4385570801 hasBestOaLocation W43855708011 @default.
- W4385570801 hasConcept C119857082 @default.
- W4385570801 hasConcept C127413603 @default.
- W4385570801 hasConcept C13280743 @default.
- W4385570801 hasConcept C134306372 @default.
- W4385570801 hasConcept C137293760 @default.
- W4385570801 hasConcept C144133560 @default.
- W4385570801 hasConcept C154945302 @default.
- W4385570801 hasConcept C162853370 @default.
- W4385570801 hasConcept C177148314 @default.
- W4385570801 hasConcept C185798385 @default.
- W4385570801 hasConcept C195324797 @default.
- W4385570801 hasConcept C201995342 @default.
- W4385570801 hasConcept C202444582 @default.
- W4385570801 hasConcept C204321447 @default.
- W4385570801 hasConcept C205649164 @default.
- W4385570801 hasConcept C2776187449 @default.
- W4385570801 hasConcept C2779439875 @default.
- W4385570801 hasConcept C2780451532 @default.
- W4385570801 hasConcept C33923547 @default.
- W4385570801 hasConcept C41008148 @default.
- W4385570801 hasConcept C86251818 @default.
- W4385570801 hasConcept C9652623 @default.
- W4385570801 hasConceptScore W4385570801C119857082 @default.
- W4385570801 hasConceptScore W4385570801C127413603 @default.
- W4385570801 hasConceptScore W4385570801C13280743 @default.
- W4385570801 hasConceptScore W4385570801C134306372 @default.
- W4385570801 hasConceptScore W4385570801C137293760 @default.
- W4385570801 hasConceptScore W4385570801C144133560 @default.
- W4385570801 hasConceptScore W4385570801C154945302 @default.
- W4385570801 hasConceptScore W4385570801C162853370 @default.
- W4385570801 hasConceptScore W4385570801C177148314 @default.
- W4385570801 hasConceptScore W4385570801C185798385 @default.
- W4385570801 hasConceptScore W4385570801C195324797 @default.
- W4385570801 hasConceptScore W4385570801C201995342 @default.
- W4385570801 hasConceptScore W4385570801C202444582 @default.
- W4385570801 hasConceptScore W4385570801C204321447 @default.
- W4385570801 hasConceptScore W4385570801C205649164 @default.
- W4385570801 hasConceptScore W4385570801C2776187449 @default.
- W4385570801 hasConceptScore W4385570801C2779439875 @default.
- W4385570801 hasConceptScore W4385570801C2780451532 @default.
- W4385570801 hasConceptScore W4385570801C33923547 @default.
- W4385570801 hasConceptScore W4385570801C41008148 @default.
- W4385570801 hasConceptScore W4385570801C86251818 @default.
- W4385570801 hasConceptScore W4385570801C9652623 @default.
- W4385570801 hasLocation W43855708011 @default.
- W4385570801 hasOpenAccess W4385570801 @default.
- W4385570801 hasPrimaryLocation W43855708011 @default.
- W4385570801 hasRelatedWork W138710363 @default.
- W4385570801 hasRelatedWork W1563618553 @default.
- W4385570801 hasRelatedWork W2587329402 @default.
- W4385570801 hasRelatedWork W2783089240 @default.
- W4385570801 hasRelatedWork W2971274815 @default.
- W4385570801 hasRelatedWork W2977842567 @default.
- W4385570801 hasRelatedWork W2988839259 @default.
- W4385570801 hasRelatedWork W3102854726 @default.
- W4385570801 hasRelatedWork W4226226396 @default.
- W4385570801 hasRelatedWork W3045475294 @default.
- W4385570801 isParatext "false" @default.
- W4385570801 isRetracted "false" @default.
- W4385570801 workType "article" @default.