List your company

Long-form Generation on PopQA (test)

0.8588AUROCSeSE
0.506240.597770.68930.78083Nov 20, 2025

Evaluation results

MethodLinks
SeSEModel=Gemini-3-Flash2025.110.85880.8119
DSEModel=Gemini-3-Flash2025.110.8480.8055
SeSEModel=DeepSeek-V3.12025.110.84680.8224
DSEModel=DeepSeek-V3.12025.110.79090.8094
P(True)Model=Gemini-3-Flash2025.110.78910.7791
P(True)Model=DeepSeek-V3.12025.110.72470.7709
PH-VUModel=Gemini-3-Flash2025.110.67530.73
PH-VUModel=DeepSeek-V3.12025.110.64760.7307
SCModel=Gemini-3-Flash2025.110.64610.7006
SCModel=DeepSeek-V3.12025.110.6210.7331
IL-VUModel=Gemini-3-Flash2025.110.58230.685
IL-VUModel=DeepSeek-V3.12025.110.51980.6394