Live Load Distribution Factors on LeapBridge Steel

Can anyone provide more information as to how LeapBridge Steel CE V21 (Version calculates the LL Distribution Factors? I have computed these by hand and I'm getting different results to those computed by the program. In Summary, I was getting > 10% difference when it comes to interior girder distribution factors for moment (both single and multiple-lane factors).

Somehow, after going through a "saved as" version of this model, the DF changed and matched hose calc'd by hand a lot closer but I cannot replicate these factors in the original LeapBridge Steel file.

Anyone know of a checkbox or simply a bug that can be causing this?