← all datasets

1,435 human-labeled utterances

Emerging
1papers using it
2026first seen

The '1,435 human-labeled utterances' dataset contains user requests that have been annotated by three human annotators, and it is used to evaluate the performance and recovery of routing accuracy in enterprise assistant models.

🔎 Find this dataset

Papers using 1,435 human-labeled utterances (1)

1,435 human-labeled utterances — datasets — time-series