List your company

Patience-Discounted Reward Evaluation on NoveltyBench

4.096UtilityQwen3-14B
1.556322.215662.8753.53434May 12, 2026

Evaluation results

MethodLinks
Qwen3-14BCalibration=Soft2026.054.096
Qwen3-14BCalibration=Hard2026.053.849
Qwen3-8BCalibration=Soft2026.053.843
GPT-OSS-20BCalibration=Base2026.053.786
Qwen3-8BCalibration=Hard2026.053.734
Qwen3-4BCalibration=Soft2026.053.559
Qwen3-14BCalibration=Base2026.053.392
Llama-3.2-3B-itCalibration=Base2026.053.385
Qwen3-8BCalibration=Base2026.053.378
Qwen3-4BCalibration=Hard2026.053.322
Llama-3.2-3B-itCalibration=Hard2026.053.183
Qwen3-1.7BCalibration=Soft2026.053.143
Llama-3.2-3B-itCalibration=Soft2026.053.126
Qwen3-1.7BCalibration=Hard2026.053.094
Gemma-3-4B-itCalibration=Soft2026.053.034
Qwen3-4BCalibration=Base2026.052.915
Gemma-3-12B-itCalibration=Soft2026.052.863
Gemma-3-27B-itCalibration=Soft2026.052.858
Qwen3-1.7BCalibration=Base2026.052.804
Gemma-3-4B-itCalibration=Hard2026.052.776
Gemma-3-12B-itCalibration=Hard2026.052.678
Qwen3-0.6BCalibration=Base2026.052.639
Gemma-3-27B-itCalibration=Hard2026.052.623
GPT-OSS-20BCalibration=Hard2026.052.615
Llama-3.2-1B-itCalibration=Base2026.052.56
Gemma-3-1B-itCalibration=Soft2026.052.555
Gemma-3-1B-itCalibration=Hard2026.052.527
Llama-3.2-1B-itCalibration=Hard2026.052.466
Llama-3.2-1B-itCalibration=Soft2026.052.388
Gemma-3-4B-itCalibration=Base2026.052.332
Gemma-3-1B-itCalibration=Base2026.052.303
Gemma-3-27B-itCalibration=Base2026.052.27
GPT-OSS-20BCalibration=Soft2026.052.179
Gemma-3-12B-itCalibration=Base2026.052.173
Qwen3-0.6BCalibration=Hard2026.052.04
Qwen3-0.6BCalibration=Soft2026.051.654