Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Mathematical Reasoning on GSM8K JSON (test)
Loading...
0.854
Structural Accuracy
RL-Struct
0.12392
0.31346
0.503
0.69254
Nov 29, 2025
Structural Accuracy
JSON Validity
Updated 5mo ago
Evaluation Results
Method
Method
Links
Structural Accuracy
JSON Validity
RL-Struct
Avg. Depth=2, learning...
2025.11
0.854
0.889
SFT
Avg. Depth=2, learning...
2025.11
0.582
0.658
Zero-shot
Avg. Depth=2, learning...
2025.11
0.152
0.352
Feedback
Search any
task
Search any
task
