List your company

Question Answering on PopQA (EM, F1, Retrieval Rate, Latency)

23EMMargin
14.26416.53218.821.068Nov 12, 2025

Evaluation results

MethodLinks
MarginBackbone=Qwen2.5-7B-Instruct, tau=0.352025.112323.112.41.761
VarianceBackbone=Qwen2.5-7B-Instruct, tau=0.402025.1122.822.918.21.847
EntropyBackbone=Qwen2.5-7B-Instruct, tau=0.802025.1122.422.312.41.761
NeverBackbone=Qwen2.5-7B-Instruct2025.112020.102.129
AlwaysBackbone=Qwen2.5-7B-Instruct2025.1114.614.61003.828