List your company
ResearchFormerly WizWand

Semantic Parsing on CFQ MCD avg

95Exact Match AccuracyDynamic Least-to-Most Prompting
11.69633.32354.9576.577Dec 1, 2021Feb 26, 2022May 24, 2022Aug 19, 2022Nov 14, 2022Feb 9, 2023May 7, 2023
Updated 8mo ago

Evaluation results

MethodLinks
Dynamic Least-to-Most PromptingAlgorithm Type=CFQ-specific2023.0595
LEARAlgorithm Type=CFQ-specific2023.0590.9
LIR+RIRBackbone=T5-3B, Algorithm Type=CFQ-specific2023.0583.8
Grounded Graph DecodingAlgorithm Type=CFQ-specific2023.0581.3
HPDAlgorithm Type=CFQ-specific2023.0569
LIR+RIRBackbone=T5-base, Algorithm Type=CFQ-specific2023.0567.8
LSPBackbone=T5-base, Algorithm Type=Domain general2023.0563.4
T5Backbone=T5-3B, Algorithm Type=Domain general2023.0549.5
T5-11B-modAlgorithm Type=CFQ-specific2023.0542.1
LSPBackbone=LSTM Encoder, Algorithm Type=Domain general2023.0541.9
LAGRAlgorithm Type=CFQ-specific2023.0534.9
T5Backbone=T5-base, Algorithm Type=Domain general2023.0534.6
Edge TransformerAlgorithm Type=Domain general2023.0524.7
Edge Transformer2021.1224.69
T5-smallsource_paper=Furrer et al. (2020)2021.1221.4
Universal Transformer2021.1221.26
Evolved Transformersource_paper=Furrer et al. (2020)2021.1220.8
Evolved TransformerAlgorithm Type=Domain general2023.0520.8
Universal Transformersource_paper=Keysers et al. (2020)2021.1218.9
Universal TransformerAlgorithm Type=Domain general2023.0518.9
TransformerAlgorithm Type=Domain general2023.0517.8
LSTMAlgorithm Type=Domain general2023.0514.9