List your company

Mathematical Reasoning on Math500 (Pass@1, #L)

80.4Pass@1Latent-GRPO
59.80865.15470.575.846Apr 30, 2026

Evaluation results

MethodLinks
Latent-GRPOBase Model=Qwen2.5-Math-7B, Reasoning Strategy=Latent Reasoning, Sampling Mode=No Sampling2026.0480.4929
GRPOBase Model=Qwen2.5-Math-7B, Reasoning Strategy=Explicit Reasoning2026.0475.152,982
SFTBase Model=Qwen2.5-Math-7B, Reasoning Strategy=Explicit Reasoning2026.0467.42,662
Latent-SFTBase Model=Qwen2.5-Math-7B, Reasoning Strategy=Latent Reasoning, Sampling Mode=No Sampling2026.0464.851,070
Soft-GRPOBase Model=Qwen2.5-Math-7B, Reasoning Strategy=Latent Reasoning, Sampling Mode=No Sampling2026.0460.61,133