List your company
ResearchFormerly WizWand

Sentiment Classification on CR (test)

92.7Mean AccuracyUniFew_meta
48.29259.82171.3582.879Jul 15, 2021Apr 11, 2022Jan 6, 2023Oct 4, 2023Jun 30, 2024Mar 27, 2025Dec 23, 2025
Updated 8mo ago

Evaluation results

MethodLinks
UniFew_metaShots=16/162021.0792.7
UniFewShots=16/162021.0792.2
UniFew_metaShots=02021.0792.1
P-TuningBackbone=RoBERTa-large, Training Samples (K)=16, Evaluation Protocol=Few-shot fine-tuning2021.0891.8
DARTBackbone=RoBERTa-large, Training Samples (K)=16, Evaluation Protocol=Few-shot fine-tuning2021.0891.8
Z-ICLModel=GPT-NeoX, Prompting=Direct, Training data access=false2022.1291.4
LM-BFFShots=16/162021.0791
Fine-tuning (full)Backbone=RoBERTa-large, Training Samples (K)=full, Evaluation Protocol=Full fine-tuning2021.0889.4
LM-BFFBackbone=RoBERTa-large, Training Samples (K)=16, Evaluation Protocol=Few-shot fine-tuning2021.0889
RETROPROMPTShot=16-shot, Source Domain=MR2022.0588.8
RETROPROMPTshot=16-shot, source_dataset=MR, domain_role=Target Domain2025.1288.8
LM-BFF (D-demo)Shot=16-shot, Source Domain=MR2022.0587.5
LM-BFF (D-demo)shot=16-shot, source_dataset=MR, domain_role=Target Domain2025.1287.5
GPT-3 in-context learningBackbone=RoBERTa-large, Training Samples (K)=16, Evaluation Protocol=In-context learning2021.0887.4
LM-BFF (man)Shot=16-shot, Source Domain=MR2022.0586.9
LM-BFF (man)shot=16-shot, source_dataset=MR, domain_role=Target Domain2025.1286.9
KPTShot=16-shot, Source Domain=MR2022.0586.7
KPTshot=16-shot, source_dataset=MR, domain_role=Target Domain2025.1286.7
ICL-goldModel=GPT-NeoX, Prompting=Channel, Training data access=true2022.1285.5
ICL-goldModel=GPT-J, Prompting=Channel, Training data access=true2022.1284.4
ICL-randomModel=GPT-J, Prompting=Channel, Training data access=true2022.1282.3
ICL-randomPrompting Method=Direct, Backbone=GPT-3, Oracle=true2022.1281
Z-ICLPrompting Method=Channel, Backbone=GPT-32022.1280.8
Z-ICLModel=GPT-J, Prompting=Channel, Training data access=false2022.1280.1
LM-BFF_manShots=02021.0779.5
Prompt-based zero-shotBackbone=RoBERTa-large, Training Samples (K)=0, Evaluation Protocol=Zero-shot2021.0879.5
ICL-goldPrompting Method=Direct, Backbone=GPT-3, Oracle=true2022.1279.5
ICL-randomModel=GPT-J, Prompting=Direct, Training data access=true2022.1279.1
Z-ICLModel=GPT-NeoX, Prompting=Channel, Training data access=false2022.1279
UniFewShots=02021.0778.8
Z-ICLModel=GPT-J, Prompting=Direct, Training data access=false2022.1278.8
ICL-goldModel=GPT-NeoX, Prompting=Direct, Training data access=true2022.1278.5
ICL-randomModel=GPT-NeoX, Prompting=Direct, Training data access=true2022.1278.5
ICL-randomModel=GPT-NeoX, Prompting=Channel, Training data access=true2022.1278.1
Random inputsModel=GPT-J, Prompting=Channel, Training data access=false2022.1277.8
No-demosPrompting Method=Channel, Backbone=GPT-32022.1276.6
Naive Z-ICLModel=GPT-NeoX, Prompting=Direct, Training data access=false2022.1276.2
Fine-tuningBackbone=RoBERTa-large, Training Samples (K)=16, Evaluation Protocol=Few-shot fine-tuning2021.0875.8
ICL-goldPrompting Method=Channel, Backbone=GPT-3, Oracle=true2022.1274.2
ICL-randomPrompting Method=Channel, Backbone=GPT-3, Oracle=true2022.1273.9
No-demosModel=GPT-J, Prompting=Channel, Training data access=false2022.1273.2
Random inputsModel=GPT-NeoX, Prompting=Direct, Training data access=false2022.1272.5
Z-ICLPrompting Method=Direct, Backbone=GPT-32022.1271.9
Random inputsModel=GPT-J, Prompting=Direct, Training data access=false2022.1271.1
ICL-goldModel=GPT-J, Prompting=Direct, Training data access=true2022.1268.7
No-demosPrompting Method=Direct, Backbone=GPT-32022.1268.4
Random inputsModel=GPT-NeoX, Prompting=Channel, Training data access=false2022.1268
Naive Z-ICLModel=GPT-J, Prompting=Direct, Training data access=false2022.1265.2
FTShot=16-shot, Source Domain=MR2022.0564.7
FTshot=16-shot, source_dataset=MR, domain_role=Target Domain2025.1264.7
Naive Z-ICLModel=GPT-NeoX, Prompting=Channel, Training data access=false2022.1262.4
Naive Z-ICLModel=GPT-J, Prompting=Channel, Training data access=false2022.1262.1
No-demosModel=GPT-NeoX, Prompting=Direct, Training data access=false2022.1261.5
No-demosModel=GPT-NeoX, Prompting=Channel, Training data access=false2022.1257.2
No-demosModel=GPT-J, Prompting=Direct, Training data access=false2022.1250.6
MajorityBackbone=RoBERTa-large, Training Samples (K)=0, Evaluation Protocol=Majority class2021.0850
Majority2022.1250
MajorityTraining data access=false2022.1250