After 10 Heads in a Row, Top Prediction for the 11th Is Heads at 82%

Written by

in

For eight base and instruction-tuned open models, the probability of predicting heads after 10 heads in a row was as high as 82%. Even when told that the coin was fair and the flips were independent, these models reduced the bias toward consecutive heads by only a few percentage points.

Source:Towards Data Science · Read Original

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *