Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Mathematical Reasoning on GSM8K (Acc, Steps)
Loading...
79.55
Accuracy
Top-1
35.662
47.056
58.45
69.844
Nov 7, 2025
Accuracy
Reasoning Steps
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy
Reasoning Steps
Top-1
Parallel=No, Backbone=...
2025.11
79.55
256
KLASS
Parallel=Yes, Backbone...
2025.11
79.43
155.67
KL divergence
Parallel=Yes, Backbone...
2025.11
76.7
150.02
KLASS
Parallel=Yes, Backbone...
2025.11
76.5
98.57
Confidence
Parallel=Yes, Backbone...
2025.11
75.21
74.35
Top-1
Parallel=No, Backbone=...
2025.11
75.13
256
KL divergence
Parallel=Yes, Backbone...
2025.11
74.52
155.88
Confidence
Parallel=Yes, Backbone...
2025.11
73.67
74.81
Top-2
Parallel=Yes, Backbone...
2025.11
72.4
128
Top-2
Parallel=Yes, Backbone...
2025.11
71.69
128
Random
Parallel=No, Backbone=...
2025.11
67.1
256
Random
Parallel=No, Backbone=...
2025.11
37.35
256
Feedback
Search any
task
Search any
task