List your company

Retrieval-Augmented Generation on TriviaQA (Reliability Score)

80.67Reliability ScoreCDA-M
45.767654.828863.8972.9512Dec 17, 2024

Evaluation results

MethodLinks
CDA-MBackbone=LLAMA3 8B INSTRUCT2024.1280.67
CDABackbone=LLAMA3 8B INSTRUCT2024.1280.62
FSBBackbone=LLAMA3 8B INSTRUCT2024.1279.14
CDA-MBackbone=LLAMA2 13B CHAT2024.1278.74
CDABackbone=LLAMA2 13B CHAT2024.1277.89
ABSTAINBackbone=LLAMA3 8B INSTRUCT2024.1276.66
ENTROPY (first-token)Backbone=LLAMA3 8B INSTRUCT2024.1276.2
ACD-ABackbone=LLAMA3 8B INSTRUCT2024.1275.58
ENTROPY (max)Backbone=LLAMA2 13B CHAT2024.1275.11
ENTROPY (first-token)Backbone=LLAMA2 13B CHAT2024.1274.56
ENTROPY (max)Backbone=LLAMA3 8B INSTRUCT2024.1274.24
ACDBackbone=LLAMA3 8B INSTRUCT2024.1273.91
CDA-MBackbone=LLAMA2 7B CHAT2024.1273.9
ENTROPY (average)Backbone=LLAMA2 13B CHAT2024.1273.88
CDA-MBackbone=MISTRAL 7B INSTRUCT2024.1273.22
CDABackbone=LLAMA2 7B CHAT2024.1273.18
FSBBackbone=LLAMA2 7B CHAT2024.1273.13
ABSTAINBackbone=MISTRAL 7B INSTRUCT2024.1273
CDABackbone=MISTRAL 7B INSTRUCT2024.1272.86
FSBBackbone=LLAMA2 13B CHAT2024.1272.79
ENTROPY (average)Backbone=LLAMA3 8B INSTRUCT2024.1272.12
ACDBackbone=LLAMA2 13B CHAT2024.1272.09
ENTROPY (max)Backbone=LLAMA2 7B CHAT2024.1271.88
ENTROPY (first-token)Backbone=LLAMA2 7B CHAT2024.1271.87
ENTROPY (first-token)Backbone=MISTRAL 7B INSTRUCT2024.1271.69
ENTROPY (max)Backbone=MISTRAL 7B INSTRUCT2024.1271.57
ENTROPY (average)Backbone=MISTRAL 7B INSTRUCT2024.1271.54
ABSTAINBackbone=LLAMA2 7B CHAT2024.1270.7
ACD-ABackbone=LLAMA2 7B CHAT2024.1270.55
ACD-ABackbone=MISTRAL 7B INSTRUCT2024.1270.39
FSBBackbone=MISTRAL 7B INSTRUCT2024.1270.26
ENTROPY (min)Backbone=LLAMA2 13B CHAT2024.1270.18
ENTROPY (average)Backbone=LLAMA2 7B CHAT2024.1270.12
ENTROPY (min)Backbone=LLAMA3 8B INSTRUCT2024.1268.61
SELF-ASKBackbone=MISTRAL 7B INSTRUCT2024.1267.94
SELF-ASKBackbone=LLAMA2 13B CHAT2024.1267.73
ACDBackbone=LLAMA2 7B CHAT2024.1267.35
CONTEXTBackbone=LLAMA2 13B CHAT2024.1266.81
ACDBackbone=MISTRAL 7B INSTRUCT2024.1266.24
ENTROPY (min)Backbone=MISTRAL 7B INSTRUCT2024.1266.19
ACD-ABackbone=LLAMA2 13B CHAT2024.1265.78
ENTROPY (min)Backbone=LLAMA2 7B CHAT2024.1265.67
CONTEXTBackbone=LLAMA3 8B INSTRUCT2024.1265.3
SELF-ASKBackbone=LLAMA3 8B INSTRUCT2024.1264.91
SELF-ASKBackbone=LLAMA2 7B CHAT2024.1263.12
CONTEXTBackbone=MISTRAL 7B INSTRUCT2024.1262.4
CADBackbone=LLAMA2 13B CHAT2024.1261.8
CONTEXTBackbone=LLAMA2 7B CHAT2024.1261.01
CADBackbone=LLAMA3 8B INSTRUCT2024.1259.32
CADBackbone=MISTRAL 7B INSTRUCT2024.1256.09
CADBackbone=LLAMA2 7B CHAT2024.1254.44
ABSTAINBackbone=LLAMA2 13B CHAT2024.1247.11