List your company

Operational decision-making on hand-crafted simulator

0.6Throughput ImprovementSFT + DPO
-2.208-1.479-0.75-0.021Mar 25, 2026

Evaluation results

MethodLinks
SFT + DPOTraining iterations=2, Base Model=Qwen2.5-14B-Instruct2026.030.60.5
SFT + DPOTraining iterations=1, Base Model=Qwen2.5-14B-Instruct2026.030.30.1
DPO onlyTraining iterations=2, Base Model=Qwen2.5-14B-Instruct2026.030-0.2
Human DecisionsProtocol=replayed2026.030—
DPO onlyTraining iterations=1, Base Model=Qwen2.5-14B-Instruct2026.03-0.4-0.7
SFT onlyBase Model=Qwen2.5-14B-Instruct2026.03-0.5-0.6
Self-RefineBase Model=Qwen 32B2026.03-2.1-2.4