sBLIMP
A zero-shot spoken language understanding evaluation task used to test the syntactic-level linguistic knowledge acquired by speech models.
Timeline (last 90 days)
On the three tasks sWuggy, sBLIMP, and tSC, GSLM2 narrows the gap between speech models and text models by about two orders of magnitude.
The sBLIMP task examines syntactic ability, requiring the model to select the grammatical sentence from a pair of grammatically correct and grammatically incorrect sentences.
All Facts (2)
On the three tasks sWuggy, sBLIMP, and tSC, GSLM2 narrows the gap between speech models and text models by about two orders of magnitude.
50%UnverifiedThe sBLIMP task examines syntactic ability, requiring the model to select the grammatical sentence from a pair of grammatically correct and grammatically incorrect sentences.
50%