Closed chojw closed 3 years ago
b is the distribution bias, which is the answer distribution under certain question types through the whole dataset. The details can be found in our paper, Sec.C in supp, and the get_bias function in main.py.
Got it thanks!
Hi,
I was just wondering, what exactly is "b" from the dataset, or entry["bias"]?
Are these the question types? I found that they are the shape of answer labels and I don't understand why they are in that shape.
Thanks in advance!