Multilingual Retrieval-Augmented Generation on MKQA 1.0 (test) benchmark leaderboard · SOTA2 Research

Multilingual Retrieval-Augmented Generation on MKQA 1.0 (test)

0.6255Accuracy (AR)DELTA

Evaluation Results

Method	Links
DELTA2026.01		0.6255	0.6385	0.6303	0.6259	0.6238	0.6286	0.6326	0.6251	0.6288	1.13
DELTA2026.01		0.5946	0.5985	0.5861	0.5967	0.5902	0.5925	0.5351	0.5645	0.5823	1.13
DELTA2026.01		0.5645	0.5697	0.5595	0.5583	0.5618	0.5598	0.5644	0.5645	0.5628	1.48
English Translation2026.01		0.5514	—	0.6194	0.5953	0.5929	0.6072	0.5457	0.6046	0.5881	1.17
English Translation2026.01		0.5097	—	0.5832	0.5611	0.5652	0.5692	0.5049	0.5752	0.5526	2.05
English Translation2026.01		0.4844	—	0.5584	0.5359	0.5368	0.5517	0.4767	0.5486	0.5275	1.55
CrossRAG2026.01		0.4834	0.6783	0.6224	0.3905	0.4927	0.6133	0.3985	0.457	0.517	2.64
MultiRAG2026.01		0.4779	0.7005	0.6376	0.3752	0.466	0.6381	0.4014	0.4073	0.513	1.38
QTT-RAG2026.01		0.4644	0.7011	0.6302	0.3768	0.4694	0.6279	0.4413	0.4212	0.5165	1.8
QTT-RAG2026.01		0.4613	0.6828	0.6181	0.3724	0.4736	0.6048	0.4106	0.4129	0.5046	1.93
DKM-RAG2026.01		0.4407	0.6784	0.6249	0.3763	0.4566	0.6165	0.403	0.4038	0.5	2.39
CrossRAG2026.01		0.4395	0.6821	0.6114	0.3781	0.4475	0.6016	0.3813	0.4287	0.4963	1.29
MultiRAG2026.01		0.4364	0.6077	0.5622	0.3314	0.4472	0.5416	0.3421	0.368	0.4546	2.56
DKM-RAG2026.01		0.4269	0.6913	0.6212	0.3513	0.439	0.6113	0.3949	0.3888	0.4906	3.8
QTT-RAG2026.01		0.4264	0.6532	0.5781	0.3156	0.4518	0.5627	0.4065	0.3597	0.4693	5.55
CrossRAG2026.01		0.4187	0.634	0.5724	0.2974	0.4414	0.568	0.3609	0.3249	0.4522	2.6
MultiRAG2026.01		0.4079	0.5826	0.5511	0.3081	0.4426	0.5352	0.3597	0.3165	0.438	1.53
DKM-RAG2026.01		0.3941	0.6421	0.5926	0.3134	0.4345	0.5774	0.3726	0.3364	0.4579	5.63