List your company

Mathematical Problem Solving on Olympiad (Pass@k)

47.1Pass@kGCPO
30.14834.54938.9543.351May 12, 2026

Evaluation results

MethodLinks
GCPOk=64, Base Model=Qwen3-1.7B2026.0547.1
GCPOk=32, Base Model=Qwen3-1.7B2026.0546.5
DQOk=64, Base Model=Qwen3-1.7B2026.0545.4
GCPOk=16, Base Model=Qwen3-1.7B2026.0545.2
GRPOk=64, Base Model=Qwen3-1.7B2026.0544.3
DQOk=32, Base Model=Qwen3-1.7B2026.0544.1
GCPOk=8, Base Model=Qwen3-1.7B2026.0543.9
Basek=64, Base Model=Qwen3-1.7B2026.0543.8
DQOk=16, Base Model=Qwen3-1.7B2026.0543.7
GRPOk=32, Base Model=Qwen3-1.7B2026.0543.6
GRPOk=16, Base Model=Qwen3-1.7B2026.0542.8
DQOk=8, Base Model=Qwen3-1.7B2026.0542.4
Basek=32, Base Model=Qwen3-1.7B2026.0542.3
GCPOk=4, Base Model=Qwen3-1.7B2026.0541.6
GRPOk=8, Base Model=Qwen3-1.7B2026.0541.1
Basek=16, Base Model=Qwen3-1.7B2026.0541
DQOk=4, Base Model=Qwen3-1.7B2026.0540.8
GRPOk=4, Base Model=Qwen3-1.7B2026.0539.4
GCPOk=2, Base Model=Qwen3-1.7B2026.0539.2
Basek=8, Base Model=Qwen3-1.7B2026.0538.3
DQOk=2, Base Model=Qwen3-1.7B2026.0537.5
GRPOk=2, Base Model=Qwen3-1.7B2026.0536
Basek=4, Base Model=Qwen3-1.7B2026.0535.1
Basek=2, Base Model=Qwen3-1.7B2026.0530.8