arXiv ScienceSearch

EXPLORE CONNECTIONS

The Belebele Benchmark: a Parallel Reading Comprehension Dataset in 122 Language Variants

Follow the relationships that help you find your next source.

Connections are not prepared for this record yet, or its source metadata has changed. The original record remains available while background snapshots are rebuilt.