List your company

Knowledge-intensive Question Answering on GPQA (Accuracy)

48.5AccuracyDIVER
19.06826.70934.3541.991May 12, 2026

Evaluation results

MethodLinks
DIVERBackbone=Qwen3-4B2026.0548.5
GCPOBackbone=Qwen3-4B2026.0547.5
DQOBackbone=Qwen3-4B2026.0547.2
DAPOBackbone=Qwen3-4B2026.0546
Div-R1Backbone=Qwen3-4B2026.0545.2
GRPOBackbone=Qwen3-4B2026.0544.4
GCPOBackbone=Qwen3-1.7B2026.0531.8
DIVERBackbone=Qwen3-1.7B2026.0530.2
DQOBackbone=Qwen3-1.7B2026.0528.3
DAPOBackbone=Qwen3-1.7B2026.0528.2
Qwen3-4B(Base)Backbone=Qwen3-4B2026.0526.3
Div-R1Backbone=Qwen3-1.7B2026.0525.1
GRPOBackbone=Qwen3-1.7B2026.0524.3
Qwen3-1.7B(Base)Backbone=Qwen3-1.7B2026.0520.2