List your company

Mathematical Reasoning on AMC (pass@1, pass@256)

58.6Pass@1CISPO
28.64836.42444.251.976Feb 6, 2026

Evaluation results

MethodLinks
CISPOBackbone=Qwen2.5-1.5B-Math, Group Size (N)=82026.0258.695.7
CISPOBackbone=Qwen2.5-7B, Group Size (N)=82026.0257.896.1
F-GRPOBackbone=Qwen2.5-7B, Group Size (N)=82026.0256.296.3
F-DAPOBackbone=Qwen2.5-7B, Group Size (N)=82026.0255.993.7
DAPOBackbone=Qwen2.5-1.5B-Math, Group Size (N)=82026.0254.595.2
F-DAPOBackbone=Qwen2.5-1.5B-Math, Group Size (N)=82026.0254.497
F-CISPOBackbone=Qwen2.5-1.5B-Math, Group Size (N)=82026.0253.899.1
DAPOBackbone=Qwen2.5-7B, Group Size (N)=82026.0253.391.9
F-CISPOBackbone=Qwen2.5-7B, Group Size (N)=82026.0253.397.1
GRPOBackbone=Qwen2.5-1.5B-Math, Group Size (N)=82026.0253.196.2
GRPOBackbone=Qwen2.5-7B, Group Size (N)=82026.0252.987.3
F-GRPOBackbone=Qwen2.5-1.5B-Math, Group Size (N)=82026.0251.695.9
F-CISPOBackbone=Llama-3.2-3B-Instruct, Group Size (N)=82026.0234.182.6
DAPOBackbone=Llama-3.2-3B-Instruct, Group Size (N)=82026.0233.179.5
CISPOBackbone=Llama-3.2-3B-Instruct, Group Size (N)=82026.0232.979.1
F-DAPOBackbone=Llama-3.2-3B-Instruct, Group Size (N)=82026.0231.988.3
GRPOBackbone=Llama-3.2-3B-Instruct, Group Size (N)=82026.0230.588.2
F-GRPOBackbone=Llama-3.2-3B-Instruct, Group Size (N)=82026.0229.890.6