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 /MATH2 /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
210 Bytes
{
"problem_id": "MATH2",
"total_energy_at_2": 132.143241,
"natural_state_lambda": 0.859642,
"ivp_outer_stretch": 1.348037,
"annular_bvp_energy": 217.135323,
"large_domain_virial_ratio": -9.819097
}