List your company

Scientific Reasoning Question Answering on QASC

74.61AccuracyS-DPO
64.854867.387469.9272.4526May 11, 2026

Evaluation results

MethodLinks
S-DPOModel=Llama32026.0574.61
DPO-kModel=Llama32026.0573.95
MASS-DPOModel=Llama32026.0573.62
MASS-DPOModel=Qwen32026.0572.19
DPO-kModel=Qwen32026.0571.96
MASS-DPOModel=SmolLM32026.0571.63
S-DPOModel=Qwen32026.0571.08
DPOModel=Llama32026.0571.08
DPO-kModel=SmolLM32026.0569.98
DMPOModel=Llama32026.0569.87
S-DPOModel=SmolLM32026.0569.43
DPOModel=Qwen32026.0568.43
DPOModel=SmolLM32026.0567
DMPOModel=Qwen32026.0566.78
DMPOModel=SmolLM32026.0565.23