List your company

General Reasoning on Average (HumanEval, GSM8K, Math-500, MT-Bench)

4.82SpeedUpPARD-2
0.84721.87862.913.9414May 9, 2026

Evaluation results

MethodLinks
PARD-2Target Model=Qwen3-8B, Inference Draft Length=16, Temperature=02026.054.826.14
PARD-2Target Model=Qwen3-14B, Inference Draft Length=16, Temperature=02026.054.796.03
PARD-2Target Model=Qwen3-32B, Inference Draft Length=16, Temperature=02026.054.685.75
PARDTarget Model=Qwen3-14B, Inference Draft Length=16, Temperature=02026.054.515.14
PARDTarget Model=Qwen3-8B, Inference Draft Length=16, Temperature=02026.054.395.56
PARDTarget Model=Qwen3-32B, Inference Draft Length=16, Temperature=02026.054.375.33
ARTarget Model=Qwen3-8B, Temperature=02026.0511
ARTarget Model=Qwen3-14B, Temperature=02026.0511
ARTarget Model=Qwen3-32B, Temperature=02026.0511