Datasets:

ulamai
/

STEM-RL-Tasks

Modalities:

Formats:

Languages:

Size:

Tags:

scientific-reasoning

reinforcement-learning

verifiable-rewards

numerical-reasoning

Libraries:

Dataset card Data Studio Files Files and versions

STEM-RL-Tasks / trajectory /BIO2 /attempt_1 /trajectory.json

Commit History

Oct 30 2 days ago

12:27
Publish STEM RL Tasks: 40 tasks across eight domains with verifiers and recorded attempts

przemekch

Verified signature
6395f35