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

wildtrace / eval /run_eval.py

Commit History

Fix evaluation harness and normalize rubrics

b37dd4f
verified

CinderD commited on about 11 hours ago

Release WildTrace strict481 public package

950f6b2
verified

CinderD commited on Jun 30