Datasets:

CinderD
/

wildtrace

Tasks:

Question Answering

Text Generation

Modalities:

Formats:

Sub-tasks:

document-question-answering

Languages:

Size:

ArXiv:

Tags:

multi-hop-reasoning

source-internal-reasoning

evidence-withheld

Libraries:

License:

Dataset card Data Studio Files Files and versions

1 contributor

History: 8 commits

CinderD's picture

Fix evaluation harness and normalize rubrics

b37dd4f verified 23 days ago