List your company

Expert Scientific Reasoning on GPQA Diamond (Accuracy, Drop, TC2b)

66.16AccuracyQwen3-32B (FP16)
15.210428.437741.66554.8923Jun 1, 2026

Evaluation results

MethodLinks
Qwen3-32B (FP16)Model=Qwen3-32B, Precision=FP162026.0666.16——
Qwen3-8B (FP16)Model=Qwen3-8B, Precision=FP162026.0659.09——
Qwen3-32B (+L)Model=Qwen3-32B, Precision=2-bit + loop rescue2026.0652.53-23.2380.81
Qwen3-8B (+P + L)Model=Qwen3-8B, Precision=2-bit + FP16 plan + loop rescue2026.0652.53-41.9231.31
Qwen3-32B (+P + L)Model=Qwen3-32B, Precision=2-bit + FP16 plan + loop rescue2026.0651.78-23.2380.81
Qwen3-32B (+P)Model=Qwen3-32B, Precision=2-bit + FP16 plan2026.0648.22-23.2380.81
Qwen3-8B (+L)Model=Qwen3-8B, Precision=2-bit + loop rescue2026.0647.98-41.9231.31
Qwen3-32B (2bit)Model=Qwen3-32B, Precision=2-bit2026.0642.93-23.2380.81
Qwen3-8B (2bit)Model=Qwen3-8B, Precision=2-bit2026.0617.17-41.9231.31
Qwen3-8B (+P)Model=Qwen3-8B, Precision=2-bit + FP16 plan2026.0617.17-41.9231.31