List your company
ResearchFormerly WizWand

Information Retrieval on Academic Search Physics

0.207MAP@100BERT-T
0.124840.146170.16750.18883Feb 16, 2026
Updated 8mo ago

Evaluation results

MethodLinks
BERT-TTraining Paradigm=Full Model, Model Architecture=BERT2026.020.2070.6570.364
DistilBERT-STraining Paradigm=Full Model, Model Architecture=DistilBERT2026.020.20.650.356
DistilBERT-HoulsbyTraining Paradigm=Knowledge Distillation (Oracle), Model Architecture=DistilBERT, Adapter Type=Houlsby2026.020.1960.6460.351
DistilBERT-HoulsbyTraining Paradigm=DRAMA, Model Architecture=DistilBERT, Adapter Type=Houlsby2026.020.1960.6470.352
DistilBERT-ALLTraining Paradigm=Full Model, Model Architecture=DistilBERT2026.020.190.6370.345
DistilBERT-LoRATraining Paradigm=Knowledge Distillation (Oracle), Model Architecture=DistilBERT, Adapter Type=LoRA2026.020.190.6380.345
DistilBERT-LoRATraining Paradigm=DRAMA, Model Architecture=DistilBERT, Adapter Type=LoRA2026.020.190.6370.345
MiniLM_SBackbone=MiniLM, Training Paradigm=Full Model, Training Scope=S2026.020.1820.6260.335
MiniLM_HoulsbyBackbone=MiniLM, Training Paradigm=Knowledge Distillation (Oracle), Adapter Type=Houlsby2026.020.1820.6260.335
MiniLM_HoulsbyBackbone=MiniLM, Training Paradigm=DRAMA, Adapter Type=Houlsby2026.020.1820.6270.335
MiniLM_LoRABackbone=MiniLM, Training Paradigm=Knowledge Distillation (Oracle), Adapter Type=LoRA2026.020.1810.6240.332
MiniLM_LoRABackbone=MiniLM, Training Paradigm=DRAMA, Adapter Type=LoRA2026.020.1810.6240.332
MiniLM_ALLBackbone=MiniLM, Training Paradigm=Full Model, Training Scope=ALL2026.020.1780.620.33
MiniLM_SArchitecture=CrossEncoder, Training Paradigm=Full Model, Model Variant=Specialized2026.020.1740.6240.328
MiniLM_HoulsbyArchitecture=CrossEncoder, Training Paradigm=Knowledge Distillation (Oracle), Adapter Type=Houlsby2026.020.1720.620.322
DRAMA (MiniLM_Houlsby)Architecture=CrossEncoder, Training Paradigm=DRAMA, Adapter Type=Houlsby2026.020.1710.6180.32
MiniLM_LoRAArchitecture=CrossEncoder, Training Paradigm=Knowledge Distillation (Oracle), Adapter Type=LoRA2026.020.1590.6050.311
DRAMA (MiniLM_LoRA)Architecture=CrossEncoder, Training Paradigm=DRAMA, Adapter Type=LoRA2026.020.1580.6040.311
MiniLM_ALLArchitecture=CrossEncoder, Training Paradigm=Full Model, Model Variant=General2026.020.1550.60.307
BM252026.020.1280.5370.269
BM25Type=Unsupervised Baseline2026.020.1280.5370.269
BM25Training Paradigm=Lexical Baseline2026.020.1280.5370.269