Evaluation logs exported from datasets/dvilasuero/arc-easy-gptoss20b-2
This dataset contains evaluation logs exported from Inspect AI.
evals: Evaluation-level data (one row per eval log)
samples: Sample-level data (one row per sample)
from datasets import load_dataset
evals = load_dataset('dvilasuero/arc-easy-gptoss20b-2', split='evals')
samples = load_dataset('dvilasuero/arc-easy-gptoss20b-2', split='samples')