Share your thoughts, 1 month free Claude Pro on usSee more

Mathematical Reasoning on GSM8K (Acc, RL, SL, RE, SE)

100Accuracy

Vanilla

Updated 2mo ago

Evaluation Results

Method	Links
Vanilla 2026.03		100	1,722	2,043	5.81	4.89
L1 2026.03		100	983	1,259	10.17	7.94
EAT 2026.03		100	1,273	1,594	7.86	6.27
ROMCSC 2026.03		100	785	1,042	12.74	9.6
ROM 2026.03		99.2	1,006	1,254	9.86	7.91
Cut2048 2026.03		98.33	1,409	2,041	6.98	4.82
Cut1024 2026.03		96.67	968	2,041	9.99	4.74
Cut512 2026.03		83.33	513	2,041	16.24	4.08