Predictions from language models for multiple-choice tasks are not robust under variation of scoring methods

Add code
Mar 01, 2024

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: