List your company

Multi-hop Reasoning on 2WikiMHQA

0.7002AUROCCoT-UQ
0.4621440.5239470.585750.647553Feb 24, 2025

Evaluation results

MethodLinks
CoT-UQModel=Llama 3.1-8B, Strategy=AP, Base Aggregation=Probas-min2025.020.7002
CoT-UQModel=Llama 2-13B, Strategy=AP, Base Aggregation=Probas-min2025.020.681
TOKENSAR + CoT-UQStrategy=AP, Backbone=Mistral 7B2025.020.6606
Probas-mean + CoT-UQStrategy=AP, Backbone=Mistral 7B2025.020.6558
CoT-UQModel=Llama 3.1-8B, Strategy=AP, Base Aggregation=TOKENSAR2025.020.6538
CoT-UQModel=Llama 3.1-8B, Strategy=AP, Base Aggregation=Probas-mean2025.020.6522
CoT-UQModel=Llama 2-13B, Strategy=AP, Base Aggregation=TOKENSAR2025.020.6403
CoT-UQModel=Llama 2-13B, Strategy=AP, Base Aggregation=Probas-mean2025.020.6329
CoT-UQModel=Llama 2-13B, Strategy=SE, Base Strategy=Self-Probing, Refinement=KEYKeywords2025.020.6263
Self-ProbingModel=Llama 2-13B, Strategy=SE2025.020.5981
CoT-UQModel=Llama 3.1-8B, Strategy=SE, Base Strategy=Self-Probing, Refinement=KEYKeywords2025.020.5838
Probas-meanStrategy=AP, Backbone=Mistral 7B2025.020.5819
CoT-UQModel=Llama 3.1-8B, Strategy=SE, Base Strategy=P(True), Refinement=KEYKeywords2025.020.5777
TOKENSARModel=Llama 3.1-8B, Strategy=AP2025.020.5692
Probas-minModel=Llama 3.1-8B, Strategy=AP2025.020.5681
Probas-meanModel=Llama 3.1-8B, Strategy=AP2025.020.568
Self-ProbingModel=Llama 3.1-8B, Strategy=SE2025.020.5639
TOKENSARStrategy=AP, Backbone=Mistral 7B2025.020.556
P(True)Model=Llama 3.1-8B, Strategy=SE2025.020.5356
CoT-UQModel=Llama 2-13B, Strategy=SE, Base Strategy=P(True), Refinement=KEYKeywords2025.020.5352
Probas-meanModel=Llama 2-13B, Strategy=AP2025.020.5154
Probas-minModel=Llama 2-13B, Strategy=AP2025.020.5128
TOKENSARModel=Llama 2-13B, Strategy=AP2025.020.5108
P(True) + CoT-UQStrategy=SE, Backbone=Mistral 7B2025.020.4978
P(True)Model=Llama 2-13B, Strategy=SE2025.020.4752
P(True)Strategy=SE, Backbone=Mistral 7B2025.020.4713