List your company

Question Answering on XQuAD

94.1F1 (de)ORACLE
59.88468.76777.6586.533Apr 30, 2020Apr 28, 2021Apr 26, 2022Apr 24, 2023Apr 21, 2024Apr 19, 2025Apr 18, 2026

Evaluation results

MethodLinks
ORACLE2026.0494.1———————92—89.8—————————91—91.7—————————————————
ORACLEStrategy=Upper bound selection, Backbone=DeepSeek-v3.12026.0494.1———————92—89.8—————————91—91.7—————————————————
TRANSLATE2026.0490.6———————88.2—82.5—————————87.2—87.1—————————————————
TRANSLATEStrategy=Translate-then-inference, Backbone=DeepSeek-v3.12026.0490.6———————88.2—82.5—————————87.2—87.1—————————————————
NATIVE2026.0489.2———————87.2—83.6—————————86.6—86.7—————————————————
XGBOOST2026.0489.2———————87.9—84.7—————————88.6—87.6—————————————————
MLP2026.0489.2———————88—83.3—————————86.5—86.7—————————————————
NATIVEStrategy=Native inference, Backbone=DeepSeek-v3.12026.0489.2———————87.2—83.6—————————86.6—86.7—————————————————
CLASSIFIER (Ours)Strategy=Learned Routing (XGBoost), Backbone=DeepSeek-v3.12026.0489.2———————87.9—84.7—————————88.6—87.6—————————————————
LTSFT2023.0680.5——73——78.6—80.6—74.3—77.8—69.7—72.2—76.5—68.9—75.9———82.4—————————————
XLMRImplementation=Standard2021.1080.486.575.7——63.479.861.78263.9————————————82.266.2————————————————
BISTIL-TFLRF=22023.0677.4——69.4——73.8—77.6—69.7—75—66.7—68.8—72.8—64.5—72.3———79.1-3.6————————————
MINILMv22023.0675.5——66.4——72.4—76.6—69.6—74—63.8—67.6—73.3—64.6—71.1———78.3-4.8————————————
XLM-R BaseBackbone=XLM-R Base, Transfer=Zero-shot from English2020.0474.483.672.166.849.160.17355.776.458.368.251.774.358.166.556.768.352.873.753.851.34270.655.5————————————————
XLM-R Base MLM-TRGBackbone=XLM-R Base, Target Language Adaptation=Masked Language Modeling (MLM), Transfer=Zero-shot from English2020.0473.784.772.66749.258.873.255.776.658.369.853.674.357.96755.868.65375.554.952.243.171.155.7————————————————
WordOTImplementation=Ours2021.1073.684.272.4——57.865.648.575.557.1————————————74.759————————————————
MAD-X BaseBackbone=XLM-R Base, Invertible Adapters=true, Transfer=Zero-shot from English2020.0472.983.572.665.548.25672.954.675.956.968.251.373.156.767.855.96749.873.753.352.742.870.354.4————————————————
MAD-XbaseImplementation=Standard2021.1072.983.572.6——5672.954.675.956.9————————————76.360————————————————
MAD-XbaseImplementation=Internal Benchmarking2021.1072.18271.1——54.571.753.774.355.7————————————7558.8————————————————
MAD-X Base-INVBackbone=XLM-R Base, Invertible Adapters=false, Transfer=Zero-shot from English2020.047283.372.16447.155.87152.974.655.567.35172.155.164.151.866.249.67353.650.940.66753.2————————————————
mBERTImplementation=Internal Benchmarking2021.107283.772.4——55.962.345.375.657.5————————————73.457.8————————————————
BISTIL-TFLRF=32023.0670.7——62.4——63.3—74.7—61.4—68.7—54.3—62.9—63—60.4—65———73.4-10.9————————————
mBERTImplementation=Standard2021.1070.683.572.2——5462.644.975.556.9————————————73.157————————————————
MINIJOINTMethod category=Proposed, Train cost (EFLOPs)=21.1, Train cost (days)=8.12022.1270.1——53.8——69.9—73.7—51.8—68.6—56.5—53.2—64.9—52.5—61.5—————————————————
BL_BASEMethod category=Standard, Train cost (EFLOPs)=54.1, Train cost (days)=20.92022.1269.2——52.4——70—74.8—53.8—68.7—57.1—57.2—65.9—53.7—62.3—————————————————
MINIPOSTMethod category=Proposed, Train cost (EFLOPs)=29.3, Train cost (days)=11.32022.1268.8——52.4——69.9—75.1—55.3—68.2—56.9—58.6—65.7—53.5—62.4—————————————————
MAD-XmBERTImplementation=Internal Benchmarking2021.1068.681.769.7——52.158.641.371.851.9————————————70.253.8————————————————
L2Implementation=Internal Benchmarking2021.1067.581.469.4——51.356.640.366.245.4————————————6851.6————————————————
XLMImplementation=Standard2021.106674.262.1——49.757.539.168.249.8————————————66.550.2————————————————
BL_SMALLMethod category=Standard, Train cost (EFLOPs)=21.1, Train cost (days)=8.12022.1261.2——48.2——63.3—65.8—46.2—61—50.4—52.1—60.3—46.2—55.5—————————————————
BL_BASE2022.12————————————————————————————2.31.11.70.83.31.92.12.21.21.5——
ByT5Scale=Small, Evaluation Protocol=Translate-train2021.05————————————————————————7459.9——————————————
ByT5Scale=Base, Evaluation Protocol=Translate-train2021.05————————————————————————78.564.6——————————————
ByT5Scale=Large, Evaluation Protocol=Translate-train2021.05————————————————————————81.467.4——————————————
ByT5Scale=XL, Evaluation Protocol=Translate-train2021.05————————————————————————83.769.5——————————————
ByT5Scale=XXL, Evaluation Protocol=Translate-train2021.05————————————————————————84.170.2——————————————
Llama_Bi_NTPmode=instruction fine-tuned2026.04————————————————————————88.719.16————————————86.78—
Llama_CLmode=instruction fine-tuned2026.04————————————————————————88.519.24————————————87.26—
Llama_CLImode=instruction fine-tuned2026.04————————————————————————85.7116.35————————————79.33—
Llama_NTPmode=instruction fine-tuned2026.04————————————————————————87.518.82————————————86.28—
Llama-3-8Bmode=instruction fine-tuned2026.04————————————————————————86.817.48————————————80.54—
MINIJOINT2022.12————————————————————————————0.80.40.60.53.10.60.6—0.50.6——
MINIPOST2022.12————————————————————————————1.40.810.41.31.31.41.20.81——
mT5Scale=Small, Evaluation Protocol=Translate-train2021.05————————————————————————71.355.7——————————————
mT5Scale=Base, Evaluation Protocol=Translate-train2021.05————————————————————————77.662.2——————————————
mT5Scale=Large, Evaluation Protocol=Translate-train2021.05————————————————————————81.366.5——————————————
mT5Scale=XL, Evaluation Protocol=Translate-train2021.05————————————————————————82.768.1——————————————
mT5Scale=XXL, Evaluation Protocol=Translate-train2021.05————————————————————————85.271.3——————————————