List your company

Cross-task decoding accuracy on SimpleQA

100Audience AccuracyQwen3-14B
4861.57588.5May 9, 2026

Evaluation results

MethodLinks
Qwen3-14BProbe training source=GSM8K / MMLU-Pro, Retraining=None2026.051001009797888194
Qwen3-30BProbe training source=GSM8K / MMLU-Pro, Retraining=None2026.051001007881725080
Qwen3-235BProbe training source=GSM8K / MMLU-Pro, Retraining=None2026.05755075751007575
Qwen3-0.6BProbe training source=GSM8K / MMLU-Pro, Retraining=None2026.0569845653697568
Llama-4Probe training source=GSM8K / MMLU-Pro, Retraining=None2026.0550505075752554