List your company

RAG Performance Prediction on HotpotQA

0.78QE5Bert-Gen
-0.074880.147060.3690.59094Apr 9, 2026

Evaluation results

MethodLinks
Bert-GenLLM=Llama, Retrieval Method=E5 Retrieval2026.040.780.7150.523
Bert-GenLLM=Llama, Retrieval Method=BM25 Retrieval2026.040.7790.6970.521
Bert-GenLLM=Falcon, Retrieval Method=E5 Retrieval2026.040.7440.6090.59
Bert-GenLLM=Falcon, Retrieval Method=BM25 Retrieval2026.040.7330.5620.599
Bert-PostLLM=Llama, Retrieval Method=E5 Retrieval2026.040.4860.2330.267
Bert-PostLLM=Llama, Retrieval Method=BM25 Retrieval2026.040.4840.2880.327
UncertaintyLLM=Llama, Retrieval Method=BM25 Retrieval2026.040.4190.0060.196
UncertaintyLLM=Llama, Retrieval Method=E5 Retrieval2026.040.385-0.0350.134
Bert-PostLLM=Falcon, Retrieval Method=BM25 Retrieval2026.040.3750.3240.405
UncertaintyLLM=Falcon, Retrieval Method=BM25 Retrieval2026.040.3710.2120.28
Bert-PostLLM=Falcon, Retrieval Method=E5 Retrieval2026.040.370.3280.378
UncertaintyLLM=Falcon, Retrieval Method=E5 Retrieval2026.040.3580.1980.245
EntailmentLLM=Llama, Retrieval Method=BM25 Retrieval2026.040.2580.1220.15
REFLLM=Llama, Retrieval Method=BM25 Retrieval2026.040.2550.1460.203
EntailmentLLM=Llama, Retrieval Method=E5 Retrieval2026.040.2330.1120.106
U_WIGLLM=Llama, Retrieval Method=BM25 Retrieval2026.040.190.1210.135
U_SMVLLM=Llama, Retrieval Method=E5 Retrieval2026.040.1810.1180.106
WIGLLM=Llama, Retrieval Method=BM25 Retrieval2026.040.180.1170.116
SMVLLM=Llama, Retrieval Method=E5 Retrieval2026.040.180.1150.105
U_WIGLLM=Llama, Retrieval Method=E5 Retrieval2026.040.1760.0650.063
REFLLM=Falcon, Retrieval Method=BM25 Retrieval2026.040.1730.1760.213
REFLLM=Llama, Retrieval Method=E5 Retrieval2026.040.1670.0990.101
QCLLM=Llama, Retrieval Method=E5 Retrieval2026.040.160.0890.069
NQCLLM=Llama, Retrieval Method=E5 Retrieval2026.040.1590.0860.068
REFLLM=Falcon, Retrieval Method=E5 Retrieval2026.040.1240.1010.143
U_WIGLLM=Falcon, Retrieval Method=BM25 Retrieval2026.040.1220.1070.143
QCLLM=Llama, Retrieval Method=BM25 Retrieval2026.040.120.0830.091
U_SMVLLM=Llama, Retrieval Method=BM25 Retrieval2026.040.120.0830.091
U_WIGLLM=Falcon, Retrieval Method=E5 Retrieval2026.040.1180.0960.144
EntailmentLLM=Falcon, Retrieval Method=BM25 Retrieval2026.040.1160.0910.173
QCLLM=Falcon, Retrieval Method=BM25 Retrieval2026.040.1120.1060.136
NQCLLM=Falcon, Retrieval Method=E5 Retrieval2026.040.1040.0870.124
EntailmentLLM=Falcon, Retrieval Method=E5 Retrieval2026.040.1030.0920.153
QCLLM=Falcon, Retrieval Method=E5 Retrieval2026.040.1020.0860.125
U_SMVLLM=Falcon, Retrieval Method=BM25 Retrieval2026.040.0910.0810.108
WIGLLM=Falcon, Retrieval Method=BM25 Retrieval2026.040.0890.0950.124
SMVLLM=Llama, Retrieval Method=BM25 Retrieval2026.040.0810.0630.054
NQCLLM=Llama, Retrieval Method=BM25 Retrieval2026.040.080.0630.054
meanSCQLLM=Llama, Retrieval Method=BM25 Retrieval2026.040.0730.0510.053
NQCLLM=Falcon, Retrieval Method=BM25 Retrieval2026.040.0670.1070.107
SMVLLM=Falcon, Retrieval Method=BM25 Retrieval2026.040.0670.1070.107
SMVLLM=Falcon, Retrieval Method=E5 Retrieval2026.040.050.0770.117
U_SMVLLM=Falcon, Retrieval Method=E5 Retrieval2026.040.0490.0760.118
maxSCQLLM=Llama, Retrieval Method=BM25 Retrieval2026.040.0440.0270.051
maxVARLLM=Llama, Retrieval Method=BM25 Retrieval2026.040.0420.0650.025
maxVARLLM=Falcon, Retrieval Method=BM25 Retrieval2026.040.040.060.057
maxSCQLLM=Falcon, Retrieval Method=BM25 Retrieval2026.040.0390.0540.058
maxIDFLLM=Falcon, Retrieval Method=BM25 Retrieval2026.040.0380.0620.048
meanSCQLLM=Llama, Retrieval Method=E5 Retrieval2026.040.0380.0370.055
minVARLLM=Llama, Retrieval Method=BM25 Retrieval2026.040.0350.030.012
maxIDFLLM=Llama, Retrieval Method=BM25 Retrieval2026.040.0310.060.047
maxSCQLLM=Llama, Retrieval Method=E5 Retrieval2026.040.030.0230.024
maxSCQLLM=Falcon, Retrieval Method=E5 Retrieval2026.040.0290.0310.039
minVARLLM=Llama, Retrieval Method=E5 Retrieval2026.040.0230.0060.029
minSCQLLM=Llama, Retrieval Method=BM25 Retrieval2026.040.020.0210.017
meanSCQLLM=Falcon, Retrieval Method=BM25 Retrieval2026.040.0190.0360.036
meanIDFLLM=Llama, Retrieval Method=BM25 Retrieval2026.040.0180.0490.033
meanVARLLM=Llama, Retrieval Method=BM25 Retrieval2026.040.0150.0530.015
minIDFLLM=Llama, Retrieval Method=BM25 Retrieval2026.040.0130.0180.013
meanSCQLLM=Falcon, Retrieval Method=E5 Retrieval2026.040.0130.0230.044
minVARLLM=Falcon, Retrieval Method=E5 Retrieval2026.040.010.010.029
WIGLLM=Llama, Retrieval Method=E5 Retrieval2026.040.008-0.012-0.018
minSCQLLM=Llama, Retrieval Method=E5 Retrieval2026.040.0070.0140.033
maxVARLLM=Llama, Retrieval Method=E5 Retrieval2026.040.0070.0340.004
maxVARLLM=Falcon, Retrieval Method=E5 Retrieval2026.040.0020.0280.012
minVARLLM=Falcon, Retrieval Method=BM25 Retrieval2026.040-0.0010.004
minIDFLLM=Llama, Retrieval Method=E5 Retrieval2026.04-0.0020.0130.025
meanVARLLM=Falcon, Retrieval Method=BM25 Retrieval2026.04-0.0070.030.016
minSCQLLM=Falcon, Retrieval Method=E5 Retrieval2026.04-0.0070.0070.024
meanIDFLLM=Falcon, Retrieval Method=BM25 Retrieval2026.04-0.0130.030.009
minSCQLLM=Falcon, Retrieval Method=BM25 Retrieval2026.04-0.013-0.0010.001
minIDFLLM=Falcon, Retrieval Method=E5 Retrieval2026.04-0.0150.0020.021
maxIDFLLM=Falcon, Retrieval Method=E5 Retrieval2026.04-0.0170.007-0.018
minIDFLLM=Falcon, Retrieval Method=BM25 Retrieval2026.04-0.022-0.007-0.001
WIGLLM=Falcon, Retrieval Method=E5 Retrieval2026.04-0.0270.0060.033
meanVARLLM=Llama, Retrieval Method=E5 Retrieval2026.04-0.0330.017-0.009
meanVARLLM=Falcon, Retrieval Method=E5 Retrieval2026.04-0.0380.007-0.01
maxIDFLLM=Llama, Retrieval Method=E5 Retrieval2026.04-0.0390.004-0.007
meanIDFLLM=Llama, Retrieval Method=E5 Retrieval2026.04-0.0410.0090.003
meanIDFLLM=Falcon, Retrieval Method=E5 Retrieval2026.04-0.042-0.001-0.024