List your company

Mathematical Reasoning on Average of Mathematical Reasoning Benchmarks

71.74AccuracyQwen3-8B + ICR
54.923259.289163.65568.0209May 8, 2026

Evaluation results

MethodLinks
Qwen3-8B + ICRBackbone=Qwen3-8B, Method=ICR2026.0571.745,133
Qwen3-4B + ICRBackbone=Qwen3-4B, Method=ICR2026.0571.334,716
Qwen3-8B + GRPOBackbone=Qwen3-8B, Method=GRPO2026.0570.645,998
Qwen3-4B + GRPOBackbone=Qwen3-4B, Method=GRPO2026.0569.455,689
Qwen3-8BBackbone=Qwen3-8B, Method=Base2026.0568.99,250
Qwen3-8B + ICR /w LP-FBackbone=Qwen3-8B, Method=ICR /w LP-F2026.0568.433,838
Qwen3-4BBackbone=Qwen3-4B, Method=Base2026.0568.48,809
Qwen3-8B + LP-FBackbone=Qwen3-8B, Method=LP-F2026.0567.624,311
Qwen3-4B + ICR /w LP-FBackbone=Qwen3-4B, Method=ICR /w LP-F2026.0567.514,101
Qwen3-4B + LP-FBackbone=Qwen3-4B, Method=LP-F2026.0566.274,712
Qwen3-8B + LC-R1Backbone=Qwen3-8B, Method=LC-R12026.0565.334,799
Qwen3-8B + ShorterBetterBackbone=Qwen3-8B, Method=ShorterBetter2026.0565.34,075
Qwen3-8B + LP-G (Kimi)Backbone=Qwen3-8B, Method=LP-G (Kimi)2026.0564.433,967
DSQW-7B + ICR /w LP-FBackbone=DSQW-7B, Method=ICR /w LP-F2026.0562.625,058
DSQW-7B + ICRBackbone=DSQW-7B, Method=ICR2026.0561.656,448
DSQW-7B + LP-FBackbone=DSQW-7B, Method=LP-F2026.0561.65,585
DSQW-7B + LC-R1Backbone=DSQW-7B, Method=LC-R12026.0561.275,940
DSQW-7B + ShorterBetterBackbone=DSQW-7B, Method=ShorterBetter2026.0561.125,144
Qwen3-4B + ShorterBetterBackbone=Qwen3-4B, Method=ShorterBetter2026.0560.883,302
DSQW-7B + GRPOBackbone=DSQW-7B, Method=GRPO2026.0560.67,384
DSQW-7B + LP-G (Kimi)Backbone=DSQW-7B, Method=LP-G (Kimi)2026.0560.585,792
Qwen3-4B + LP-G (Kimi)Backbone=Qwen3-4B, Method=LP-G (Kimi)2026.0560.332,993
Qwen3-4B + LC-R1Backbone=Qwen3-4B, Method=LC-R12026.0560.184,420
DSQW-7BBackbone=DSQW-7B, Method=Base2026.0555.577,624