List your company

Question Answering on GPQA Think (test)

61.1AccuracyQwen3-8B
46.74850.47454.257.926Jan 11, 2026

Evaluation results

MethodLinks
Qwen3-8BTraining=RL, Mode=Mid-Think2026.0161.18,25755,272
Qwen3-8BTraining=No, Mode=Untrained2026.0160.99,00871,817
Qwen3-8BTraining=RL, Mode=No-Think2026.0159.69,04759,419
Qwen3-8BTraining=RL, Mode=Think2026.0158.59,41352,228
Qwen3-4BTraining=RL, Mode=Mid-Think2026.0155.38,69660,790
Qwen3-4BTraining=No, Mode=Untrained2026.0153.28,72972,509
Qwen3-4BTraining=RL, Mode=No-Think2026.0148.710,77482,568
Qwen3-4BTraining=RL, Mode=Think2026.0147.310,48158,476