II.
TestSet overview
Reference · livetest-set:math-test
MATH test split overview
Canonical MATH test set. Often subsetted ("MATH-500") in modern leaderboards; this entry tracks the full 5k split.
Attributes
displayName
MATH test split
benchmarkId
caseCount
5000
releasedAt
2021-03-05
composition
The 5,000-problem test split of the MATH benchmark (Hendrycks et
al., 2021). Problems span seven subjects (algebra, geometry,
number theory, counting & probability, intermediate algebra,
pre-algebra, pre-calculus) and five difficulty levels.
homepageUrl
description
Canonical MATH test set. Often subsetted ("MATH-500") in modern
leaderboards; this entry tracks the full 5k split.
Outgoing edges
belongs_to_benchmark1
- benchmark:math·BenchmarkMATH
Incoming edges
None.