List your company

Mathematical Reasoning on MATH (MV, EM, PK)

89.4Pass@k (PK)Forking
39.06452.13265.278.268Apr 2, 2026Apr 6, 2026Apr 11, 2026Apr 16, 2026Apr 20, 2026Apr 25, 2026Apr 30, 2026

Evaluation results

MethodLinks
ForkingBackbone=Qwen2.5-Math-7B, Method Category=Policy Entropy Method2026.0489.4——84.1
EAPOBackbone=Qwen2.5-Math-7B, Method Category=Policy Entropy Method2026.0488.7——83.7
EntroAdvBackbone=Qwen2.5-Math-7B, Method Category=Policy Entropy Method2026.0488.6——83.5
DAPOBackbone=Qwen2.5-Math-7B, Method Category=Classic RLVR Method2026.0488.1——82.6
EntroRegBackbone=Qwen2.5-Math-7B, Method Category=Policy Entropy Method2026.0487.7——82.7
KL-CovBackbone=Qwen2.5-Math-7B, Method Category=Policy Entropy Method2026.0487.2——81.5
GRPOBackbone=Qwen2.5-Math-7B, Method Category=Classic RLVR Method2026.0485.1——79.9
W-REINFBackbone=Qwen2.5-Math-7B, Method Category=Classic RLVR Method2026.0484.5——76.2
I-PPOBackbone=Qwen2.5 3B2026.0479.666.8551.62—
PPOBackbone=Qwen2.5 3B2026.0476.3961.2243.84—
PRIMEBackbone=Qwen2.5-Math-7B, Method Category=Classic RLVR Method2026.0474.1——56.6
p* smooth(alpha = 0.01)Setting=ZVF2026.0473.3———
No samplingSetting=GRPO2026.0473.2———
p*Setting=GRPO2026.0473.2———
No samplingSetting=ZVF2026.0473.2———
p*Setting=ZVF2026.0472.9———
I-PPOBackbone=LLaMa3 8B2026.0472.855.2543.42—
PPOBackbone=LLaMa3 8B2026.0472.7752.0838.71—
p* smooth(alpha = 0.01)Setting=GRPO2026.0472.3———
BaselineBackbone=Qwen2.5 3B2026.0471.551.735.82—
BaselineBackbone=LLaMa3 8B2026.0468.545.128.79—
I-PPOBackbone=Phi3 4B2026.0467.552.9146.46—
Base-7BBackbone=Qwen2.5-Math-7B, Method Category=Classic RLVR Method2026.0466.3——46.9
PPOBackbone=Phi3 4B2026.0465.449.8641.4—
BaselineBackbone=Phi3 4B2026.0459.338.2531.11—
I-PPOBackbone=Gemma2 2B2026.0459.145.0738.71—
PPOBackbone=Gemma2 2B2026.0455.5640.6629.36—
BaselineBackbone=Gemma2 2B2026.045133.520.82—
I-PPOBackbone=Rho 1B2026.0449.1423.819.01—
PPOBackbone=Rho 1B2026.0447.6221.415.33—
BaselineBackbone=Rho 1B2026.044112.28.41—
DeepSeek-V3.2 Exp-Baseshots=4-shot2026.06——60.12—
GLM-4.5 Baseshots=4-shot2026.06——57.58—
Kimi-K2 Baseshots=4-shot2026.06——68.4—
Mistral-Large-3 675B-Base-2512shots=4-shot2026.06——62.88—
Nemotron-3-Ultra 550B-A55B-Baseshots=4-shot2026.06——82—