List your company

Mathematical Reasoning on AIME 25 (pass@16)

36.2pass@16P-trace-0.9
24.021627.183330.34533.5067May 7, 2026

Evaluation results

MethodLinks
P-trace-0.9Backbone=Qwen3-4B2026.0536.2
S-trace-0.9Backbone=Qwen3-4B2026.0532.18
OPOBackbone=Qwen3-4B2026.0532.06
GRPOBackbone=Qwen3-4B2026.0526.67
GRPO(λ)-0.9Backbone=Qwen3-4B2026.0526.66
GRPO(λ)-0.99Backbone=Qwen3-4B2026.0524.49