List your company

Document Retrieval on Russian-Wikipedia corpus 10^6 documents

0.834Acc@1Baseline_proj (SVD k)
0.64160.691550.74150.79145Jun 24, 2026

Evaluation results

MethodLinks
Baseline_proj (SVD k)Encoder=multilingual-e5-base, Original Dimension (d)=768, Projected Dimension (k)=384, Subspaces (M)=96, sigma_rec=0.1292026.060.8340.9240.8650.879——
ProposedEncoder=multilingual-e5-base, Original Dimension (d)=768, Projected Dimension (k)=384, Subspaces (M)=96, sigma_rec=0.1292026.060.8330.9240.8640.8790.003356
Baseline_proj (SVD k)Encoder=BAAI/bge-m3, Original Dimension (d)=1024, Projected Dimension (k)=512, Subspaces (M)=128, sigma_rec=0.1282026.060.8320.9260.8630.878——
ProposedEncoder=BAAI/bge-m3, Original Dimension (d)=1024, Projected Dimension (k)=512, Subspaces (M)=128, sigma_rec=0.1282026.060.8320.9240.8630.8780.014221
Baseline (raw d)Encoder=multilingual-e5-base, Original Dimension (d)=768, Projected Dimension (k)=384, Subspaces (M)=96, sigma_rec=0.1292026.060.830.9340.8620.88——
Baseline (raw d)Encoder=BAAI/bge-m3, Original Dimension (d)=1024, Projected Dimension (k)=512, Subspaces (M)=128, sigma_rec=0.1282026.060.8180.9080.8480.862——
Baseline_proj (SVD k)Encoder=multilingual-e5-large, Original Dimension (d)=1024, Projected Dimension (k)=512, Subspaces (M)=128, sigma_rec=0.1012026.060.8140.9220.8480.866——
ProposedEncoder=multilingual-e5-large, Original Dimension (d)=1024, Projected Dimension (k)=512, Subspaces (M)=128, sigma_rec=0.1012026.060.8140.920.8480.8650.018219
Baseline (raw d)Encoder=multilingual-e5-large, Original Dimension (d)=1024, Projected Dimension (k)=512, Subspaces (M)=128, sigma_rec=0.1012026.060.7960.880.8240.838——
Baseline (raw d)Encoder=multilingual-e5-small, Original Dimension (d)=384, Projected Dimension (k)=192, Subspaces (M)=48, sigma_rec=0.2392026.060.7820.890.8170.834——
Baseline (raw d)Encoder=paraphrase-multilingual-mpnet-base-v2, Original Dimension (d)=768, Projected Dimension (k)=384, Subspaces (M)=96, sigma_rec=0.1072026.060.7160.8420.7620.782——
Baseline_proj (SVD k)Encoder=multilingual-e5-small, Original Dimension (d)=384, Projected Dimension (k)=192, Subspaces (M)=48, sigma_rec=0.2392026.060.7020.8340.7470.768——
ProposedEncoder=multilingual-e5-small, Original Dimension (d)=384, Projected Dimension (k)=192, Subspaces (M)=48, sigma_rec=0.2392026.060.7010.8310.7460.7670.081314
Baseline_proj (SVD k)Encoder=paraphrase-multilingual-mpnet-base-v2, Original Dimension (d)=768, Projected Dimension (k)=384, Subspaces (M)=96, sigma_rec=0.1072026.060.650.7960.6930.717——
ProposedEncoder=paraphrase-multilingual-mpnet-base-v2, Original Dimension (d)=768, Projected Dimension (k)=384, Subspaces (M)=96, sigma_rec=0.1072026.060.6490.7950.6920.7170.067351