GSM8K
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
GSM8K
93Accuracy
13
GSM8K (test)
97.8Accuracy
13
GSM8K ProcessBench (test)
24.2Error Rate
13
GSM8K (test)
2.66Speedup (x)
13
GSM8K Python
78.5Pass@100
13
GSM8K
83.04Accuracy@1
12
GSM8K
79.8Accuracy
12
GSM8K (test)
66.26GSM8K EM
12
GSM8K
128Samples Required
12
GSM8K ⊕ HEx-PHI
76.1Utility
12
GSM8K
92.65Accuracy
12
GSM8K
88.02Score
12
GSM8k-ZH
95.9Score
12
GSM8K
501Response Length (Tokens)
12
GSM8K
83.85Avg@8
12
GSM8K
83.85Average Score @8
12
GSM8K
83.5Avg@8
12
GSM8K
83.9Accuracy
12
GSM8K
0Average False Positive Rate
12
GSM8K
89.9Accuracy
12
GSM8K (held-out)
81.2Accuracy
12
GSM8K Standard (Full)
71.9Average Score @16
12
GSM8K
94.9Accuracy
12
GSM8K
0.16Wall-clock Time (s)
12
GSM8K 0-shot CoT
75.13Accuracy
12