STEB is a new benchmark of 96 datasets in 7 languages for evaluating style text embeddings on authorship, detection, and linguistic probing tasks.
Dongyeop Kang, Varun Gangal, and Eduard Hovy
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
STEB: Style Text Embedding Benchmark
STEB is a new benchmark of 96 datasets in 7 languages for evaluating style text embeddings on authorship, detection, and linguistic probing tasks.