List your company

Multimodal Numerical Reasoning on GeoQA (test)

92.3Total AccuracyHuman
30.4246.48562.5578.615May 30, 2021

Evaluation results

MethodLinks
HumanInput modality=Text-Diagram2021.0592.394.290.587
HumanInput modality=Text-Only2021.05635871.755.6
NGS AuxiliaryInput modality=Text-Diagram, Program-based reasoning=true, Auxiliary supervision=true2021.056071.548.829.6
NGSInput modality=Text-Diagram, Program-based reasoning=true, Auxiliary supervision=false2021.0557.468.643.544.4
BERT2ProgInput modality=Text-Only, Program-based reasoning=true2021.0554.465.74137
Seq2ProgInput modality=Text-Only, Program-based reasoning=true2021.0554.266.438.542.6
Seq2Prog + DiagramInput modality=Text-Diagram, Program-based reasoning=true2021.0553.462.444.531.5
BERT2Prog + DiagramInput modality=Text-Diagram, Program-based reasoning=true2021.0552.565.737.131.5
MCANProgram-based reasoning=false, Input modality=Text-Diagram2021.0539.543.23629.6
RNProgram-based reasoning=false, Input modality=Text-Diagram2021.0538.242.234.327.8
FiLMProgram-based reasoning=false, Input modality=Text-Diagram2021.0532.833.632.925.9