SocialIQA is the first large-scale benchmark with 38k crowdsourced questions testing commonsense about social interactions, where pretrained language models trail humans by over 20% but transfer to improve performance on Winograd Schemas and COPA.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CL 2verdicts
UNVERDICTED 2representative citing papers
Scene Abstraction framework builds structured scene representations for lexical meaning via LLM prompting, with COCA-Scenes dataset and human experiments showing 82.4% identification accuracy and 86.4% preference over ATOMIC baselines.
citing papers explorer
-
SocialIQA: Commonsense Reasoning about Social Interactions
SocialIQA is the first large-scale benchmark with 38k crowdsourced questions testing commonsense about social interactions, where pretrained language models trail humans by over 20% but transfer to improve performance on Winograd Schemas and COPA.
-
Scene Abstraction for Lexical Semantics: Structured Representations of Situated Meaning
Scene Abstraction framework builds structured scene representations for lexical meaning via LLM prompting, with COCA-Scenes dataset and human experiments showing 82.4% identification accuracy and 86.4% preference over ATOMIC baselines.