List your company
ResearchFormerly WizWand

Text Classification on Yelp.P (test)

98.63AccuracyXLNet
79.046884.130989.21594.2991Oct 22, 2018Oct 6, 2019Sep 20, 2020Sep 5, 2021Aug 20, 2022Aug 5, 2023Jul 20, 2024
Updated 8mo ago

Evaluation results

MethodLinks
XLNetfine-tuning=true2018.1098.63
BERT-Largefine-tuning=true2018.1098.11
ULMFiTfine-tuning=true2018.1097.84
BERT-FiTfine-tuning=true2018.1097.72
DPCNNfine-tuning=true2018.1097.36
CCCapsNetfine-tuning=false, trained from scratch=true2018.1096.48
Region Emb.fine-tuning=false, trained from scratch=true2018.1096.4
PINFew-shot=5-shot, Policy LM=OPT-125M2024.0795
fastTextModel=RNN2023.0294.7
RETVecModel=RNN2023.0294.7
RLPromptFew-shot=5-shot, Policy LM=OPT-125M2024.0794.7
WhitespaceModel=CNN2023.0294.4
RETVec-rawModel=RNN2023.0294.1
WhitespaceModel=RNN2023.0293.9
fastTextModel=CNN2023.0293.9
SentencePieceModel=CNN2023.0293.8
fastTextModel=BERT2023.0293.5
BPEModel=RNN2023.0293.4
RETVecModel=CNN2023.0293.4
SentencePieceModel=RNN2023.0293.3
RETVec-rawModel=BERT2023.0293.2
BPEModel=CNN2023.0293
RETVecModel=BERT2023.0292.8
RETVec-rawModel=CNN2023.0292.5
WhitespaceModel=BERT2023.0291.6
BPEModel=BERT2023.0291.6
SentencePieceModel=BERT2023.0291.3
In-Context DemonstrationFew-shot=5-shot2024.0789.6
Soft Prompt TuningFew-shot=5-shot2024.0788.6
GrIPSFew-shot=5-shot2024.0788.2
PEZFew-shot=5-shot2024.0785.9
InstructionsFew-shot=5-shot2024.0784.4
Manual PromptFew-shot=5-shot2024.0783
AutoPromptFew-shot=5-shot2024.0779.8