List your company

Mathematical Reasoning on AIME24 (pass@16)

71.7Pass@16DASD
10.75626.57842.458.222May 7, 2026May 9, 2026May 11, 2026May 14, 2026May 16, 2026May 18, 2026May 21, 2026

Evaluation results

MethodLinks
DASDModel Scale=8B2026.0571.7
HEPOModel Scale=8B2026.0566.7
RLSDModel Scale=8B2026.0562.3
GRPOModel Scale=8B2026.0559.4
DASDModel Scale=4B2026.0558.3
RLSDModel Scale=4B2026.0556.5
GRPOModel Scale=4B2026.0555
HEPOModel Scale=4B2026.0553.3
SRPOModel Scale=8B2026.0543.8
SRPOModel Scale=4B2026.0540
GRPO(λ)-0.9Backbone=Qwen3-4B2026.0538.83
RLSDModel Scale=1.7B2026.0535.4
DASDModel Scale=1.7B2026.0535
GRPOBackbone=Qwen3-4B2026.0533.33
GRPOModel Scale=1.7B2026.0532.5
HEPOModel Scale=1.7B2026.0532.5
S-trace-0.9Backbone=Qwen3-4B2026.0530
Qwen3Model Scale=8B2026.0527.9
SDPOModel Scale=8B2026.0527.9
OPSDModel Scale=8B2026.0527.5
GRPO(λ)-0.99Backbone=Qwen3-4B2026.0526.67
OPOBackbone=Qwen3-4B2026.0525.48
SDPOModel Scale=4B2026.0525
OPSDModel Scale=4B2026.0524.2
P-trace-0.9Backbone=Qwen3-4B2026.0523.33
SRPOModel Scale=1.7B2026.0523.3
Qwen3Model Scale=4B2026.0523.1
OPSDModel Scale=1.7B2026.0515
SDPOModel Scale=1.7B2026.0514.2
Qwen3Model Scale=1.7B2026.0513.1