Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Mathematical Reasoning on GSM8K (Accuracy, Friedman Rank, P(EDEN))
Loading...
81.5
Accuracy
Beam Search
69.228
72.414
75.6
78.786
May 10, 2026
Accuracy
Friedman Rank
P(EDEN Win Probability)
Updated 22d ago
Evaluation Results
Method
Method
Links
Accuracy
Friedman Rank
P(EDEN Win Probability)
Beam Search
beam width=3, total nu...
2026.05
81.5
2.25
0.78
EDEN
total number of branch...
2026.05
80.5
1.25
0.75
Majority-n
n=5
2026.05
78.7
6.25
0.99
Best-of-n
n=5
2026.05
78.2
4.62
0.96
Diverse Beam Search
beam width=3, total nu...
2026.05
77.5
4.75
0.96
Greedy
2026.05
73.5
6.12
0.99
Top-p Sampling
2026.05
73.5
6.62
0.99
Min-p Sampling
2026.05
72.3
8
1
Top-k Sampling
2026.05
70.7
7
1
Top-H Sampling
2026.05
69.7
8.12
1
Feedback
Search any
task
Search any
task