Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Mathematical Reasoning on GSM8k (Accuracy, Loss)
Loading...
59
Accuracy
Adam+Nexus
16.36
27.43
38.5
49.57
Apr 10, 2026
Accuracy
Loss
Updated 6d ago
Evaluation Results
Method
Method
Links
Accuracy
Loss
Adam+Nexus
Optimizer=Adam+Nexus,...
2026.04
59
1.227
Nexus
Model Size=3B, Optimiz...
2026.04
59
1.227
Nexus
Model scale=3B
2026.04
47
1.533
Muon
Optimizer=Muon, Model...
2026.04
46
1.236
AdamW
Optimizer=AdamW, Model...
2026.04
44
1.259
Adam
Model scale=3B
2026.04
44
1.519
AdamW
Model Size=3B, Optimiz...
2026.04
44
1.259
Nexus
Model scale=1B
2026.04
22
1.749
Nexus
Model Size=1B, Optimiz...
2026.04
20
1.396
Adam
Model scale=1B
2026.04
18
1.708
AdamW
Model Size=1B, Optimiz...
2026.04
18
1.429
Feedback
Search any
task
Search any
task