List your company

Question Answering on Science Q&A Chemistry (test)

81.5Mean@16SC-SDPO
19.72435.76251.867.838May 26, 2026

Evaluation results

MethodLinks
SC-SDPOBackbone=OLMo-3-7B, Training Method=SC-SDPO, Alpha (α)=0.5, Sampling Count (k)=162026.0581.582.5
SDPOBackbone=OLMo-3-7B, Training Method=SDPO, Sampling Count (k)=162026.0581.181.4
OLMo-3-7BBackbone=OLMo-3-7B, Training Method=Pre-trained, Sampling Count (k)=162026.0522.19.5