Skip to Main Content
Table 6: 

Results on datasets used in Gururangan et al. (2020). For vanilla (unregularized) RoBERTa, DAPT, and TAPT, results are taken from Gururangan et al. (2020). For each method on each dataset, we run it for four times with different random seeds. Results are in ms format, where m denotes mean and s denotes standard derivation. Following Gururangan et al. (2020), for ChemProt and RCT, we report micro-F1; for other datasets, we report macro-F1.

DatasetRoBERTaDAPTTAPTSSL-RegTAPT+SSL-RegDAPT+SSL-Reg
ChemProt 81.91.0 84.20.2 82.60.4 83.10.5 83.50.1 84.40.3  
RCT 87.20.1 87.60.1 87.70.1 87.40.1 87.70.1 87.70.1 
 
ACL-ARC 63.05.8 75.42.5 67.41.8 69.34.9 68.12.0 75.71.4 
SciERC 77.31.9 80.81.5 79.31.5 81.40.8 80.40.6 82.30.8 
 
HyperPartisan 86.60.9 88.25.9 90.45.2 92.31.4 93.21.8 90.73.2 
AGNews 93.90.2 93.90.2 94.50.1 94.20.1 94.40.1 94.00.1 
 
Helpfulness 65.13.4 66.51.4 68.51.9 69.40.2 71.01.0 68.31.4 
IMDB 95.00.2 95.40.1 95.50.1 95.70.1 96.10.1 95.40.1 
DatasetRoBERTaDAPTTAPTSSL-RegTAPT+SSL-RegDAPT+SSL-Reg
ChemProt 81.91.0 84.20.2 82.60.4 83.10.5 83.50.1 84.40.3  
RCT 87.20.1 87.60.1 87.70.1 87.40.1 87.70.1 87.70.1 
 
ACL-ARC 63.05.8 75.42.5 67.41.8 69.34.9 68.12.0 75.71.4 
SciERC 77.31.9 80.81.5 79.31.5 81.40.8 80.40.6 82.30.8 
 
HyperPartisan 86.60.9 88.25.9 90.45.2 92.31.4 93.21.8 90.73.2 
AGNews 93.90.2 93.90.2 94.50.1 94.20.1 94.40.1 94.00.1 
 
Helpfulness 65.13.4 66.51.4 68.51.9 69.40.2 71.01.0 68.31.4 
IMDB 95.00.2 95.40.1 95.50.1 95.70.1 96.10.1 95.40.1 
Close Modal

or Create an Account

Close Modal
Close Modal