no code implementations • 13 Apr 2023 • Swapnil Sharma, Nikita Anand, Kranthi Kiran G. V., Alind Jain
The results are compiled and we find that all the models under study do exhibit biases but have generally improved as compared to BERT.