List your company

STEM Reasoning on GPQA (Pass@1)

77.06Pass@1 AccuracyGPT-OSS-120B
25.517638.898852.2865.6612Apr 10, 2026Apr 22, 2026May 4, 2026May 17, 2026May 29, 2026Jun 10, 2026Jun 23, 2026

Evaluation results

MethodLinks
GPT-OSS-120BSampling Strategy=4-sample mean2026.0477.06
GPT-5 MiniSampling Strategy=4-sample mean2026.0475.46
Gemini 2.5 FlashSampling Strategy=4-sample mean2026.0475.09
Aryabhata 2Sampling Strategy=4-sample mean2026.0474.86
Qwen3-30B-A3B (Thinking)Sampling Strategy=4-sample mean2026.0473.31
GPT-OSS-20BSampling Strategy=4-sample mean2026.0470.51
Nemotron 3 Nano 30B A3BSampling Strategy=4-sample mean2026.0465.38
GPT-5 NanoSampling Strategy=4-sample mean2026.0461.11
TACBackbone=Qwen3-4B-Base, Seeds=12026.0642.5
RandomBackbone=Qwen3-4B-Base, Seeds=12026.0641.3
SECBackbone=Qwen3-4B-Base, Seeds=12026.0636.7
SECBackbone=Qwen3-0.6B-Base, Seeds=12026.0629.5
RandomBackbone=Qwen3-0.6B-Base, Seeds=12026.0627.7
TACBackbone=Qwen3-0.6B-Base, Seeds=12026.0627.5