Hugging Face's logo

Datasets:
ulamai
/
STEM-RL-Tasks

Modalities:
Tabular
Text
Formats:
json
Languages:
English
Size:
< 1K
Tags:
stem
scientific-reasoning
reinforcement-learning
rlvr
verifiable-rewards
numerical-reasoning
Libraries:
Datasets
pandas
Polars
Dataset card Data Studio Files
xet
 Community
1
STEM-RL-Tasks / tasks /MATH3 /solution /solution.json
przemekch's picture
przemekch
Publish STEM RL Tasks: 40 tasks across eight domains with verifiers and recorded attempts
6395f35 verified
Raw History Blame Contribute Delete
182 Bytes
{
"problem_id": "MATH3",
"complex_alpha_d1": 0.886227,
"complex_alpha_d2": 0.861653,
"complex_alpha_d3": 0.855437,
"real_alpha_d1": 0.797885,
"real_alpha_d2": 0.810201
}