List your company

Hallucination Detection on LogicStruct

73.89Neg. ScorePRISM-SAPLMA
39.767648.626357.48566.3437Nov 7, 2024

Evaluation results

MethodLinks
PRISM-SAPLMABackbone=LLaMA2-13B-Chat2024.1173.8981.1257.2270.74
PRISM-SAPLMABackbone=LLaMA2-7B-Chat2024.1173.4575.2953.2967.34
PRISM-MMBackbone=LLaMA2-7B-Chat2024.1155.4880.8755.864.05
PRISM-MMBackbone=LLaMA2-13B-Chat2024.1152.687.3462.0767.34
MINDBackbone=LLaMA2-7B-Chat2024.1152.1943.1849.6948.35
SAPLMABackbone=LLaMA2-13B-Chat2024.1151.5567.8851.8657.1
MMBackbone=LLaMA2-13B-Chat2024.1150.7148.9550.249.95
MMBackbone=LLaMA2-7B-Chat2024.1149.4849.3750.1349.66
SAPLMABackbone=LLaMA2-7B-Chat2024.1148.6451.9650.350.3
MINDBackbone=LLaMA2-13B-Chat2024.1148.4948.7350.3849.2
EUBIDBackbone=LLaMA2-13B-Chat2024.1145.1252.952.6350.22
LN-PPBackbone=LLaMA2-13B-Chat2024.1142.6352.4358.6351.23
EUBIDBackbone=LLaMA2-7B-Chat2024.1141.8453.153.8749.6
LN-PPBackbone=LLaMA2-7B-Chat2024.1141.0852.7857.4350.43