Combibench
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Combibench
48Solve Rate
15
CombiBench
8Proof Length
13
CombiBench Combinatorics
27Solved Problems
13
CombiBench
97Pass@8
13
CombiBench N = 100
1CH@100
11
CombiBench
16pass@32
8
CombiBench Easy Mode
10Solved Problems (Pass@32)
4
CombiBench (N=100)
96Pass@64
4
CombiBench Hard Mode
10Total Solved (Pass@32)
3