List your company

Mathematical Reasoning on AIME 25 (Avg@32)

36.35Avg@32OISD
-1.418648.3866818.19227.99732Oct 11, 2025Nov 22, 2025Jan 3, 2026Feb 14, 2026Mar 28, 2026May 9, 2026Jun 21, 2026

Evaluation results

MethodLinks
OISDBackbone=Qwen3-4B2026.0536.35
GRPOBackbone=Qwen3-4B2026.0528.85
OPSDBackbone=Qwen3-4B2026.0525.62
MHPOBackbone=Qwen2.5-Math-7B-Instruct2026.0321.6
MHPOBackbone=Qwen2.5-Math-7B-Instruct2026.0321.6
MHPOBackbone=Qwen2.5-7B-Instruct2026.0316.7
STEERBackbone=Qwen2.5-Math-7B2025.1016.1
DAPOBackbone=Qwen2.5-Math-7B-Instruct2026.0315
DAPOBackbone=Qwen2.5-Math-7B-Instruct2026.0315
GRPOBackbone=Qwen2.5-Math-7B2025.1014.3
GRPO w/ Fork TokensBackbone=Qwen2.5-Math-7B2025.1014.3
GSPOBackbone=Qwen2.5-Math-7B-Instruct2026.0314.2
GSPOBackbone=Qwen2.5-Math-7B-Instruct2026.0314.2
SimpleRL-ZooBackbone=Qwen2.5-Math-7B2025.1014.2
KL–CovBackbone=Qwen2.5-Math-7B2025.1014.1
GRPO w/ Entro. LossBackbone=Qwen2.5-Math-7B2025.1014
Entro. Adv.Backbone=Qwen2.5-Math-7B2025.1013.5
OPOBackbone=Qwen2.5-Math-7B2025.1013.4
SAPOBackbone=Qwen2.5-Math-7B-Instruct2026.0313.1
SAPOBackbone=Qwen2.5-Math-7B-Instruct2026.0313.1
Eurus-PRIMEBackbone=Qwen2.5-Math-7B2025.1013
Clip–CovBackbone=Qwen2.5-Math-7B2025.1012.9
GRPO w/ clip-highBackbone=Qwen2.5-Math-7B2025.1012.8
W-REINFORCEBackbone=Qwen2.5-Math-7B2025.1012
DAPOBackbone=Qwen2.5-7B-Instruct2026.0311.4
GRPOBackbone=Qwen2.5-Math-7B-Instruct2026.0311
GRPOBackbone=Qwen2.5-Math-7B-Instruct2026.0311
GSPOBackbone=Qwen2.5-7B-Instruct2026.0310.8
SAPOBackbone=Qwen2.5-7B-Instruct2026.0310.1
GRPOBackbone=Qwen2.5-7B-Instruct2026.038.6
BaselineBackbone=Qwen2.5-Math-7B-Instruct2026.037.7
BaselineBackbone=Qwen2.5-Math-7B-Instruct2026.037.7
Qwen2.5-Math-7BBackbone=Qwen2.5-Math-7B2025.105.3
BaselineBackbone=Qwen2.5-7B-Instruct2026.033.5
Qwen3-30B-A3B-Instruct-2507Role=Teacher Model2026.060.628
IW-OPDStudent=Qwen3-4B2026.060.497
IW-OPDStudent=Qwen3-4B, Training Step=102026.060.493
OPDStudent=Qwen3-4B2026.060.48
OPDStudent=Qwen3-4B, Training Step=102026.060.424
IW-OPDStudent=Qwen3-1.7B2026.060.295
OPDStudent=Qwen3-1.7B2026.060.287
IW-OPDStudent=Qwen3-1.7B, Training Step=102026.060.232
BaseStudent=Qwen3-4B2026.060.214
OPDStudent=Qwen3-1.7B, Training Step=102026.060.202
IW-OPDStudent=Qwen3-0.6B2026.060.193
OPDStudent=Qwen3-0.6B2026.060.178
IW-OPDStudent=Qwen3-0.6B, Training Step=102026.060.158
OPDStudent=Qwen3-0.6B, Training Step=102026.060.141
BaseStudent=Qwen3-1.7B2026.060.11
BaseStudent=Qwen3-0.6B2026.060.034