The results of DNA sequence and reverse complementary sequence are inconsistent

[2SSSCN]

Why is the result of DNA sequence and its complementary sequence of the model I trained inconsistent？ For example, the Score 1 of a DNA sequence is larger than 0.5 but the score 1 of its complementary sequence is smaller than 0.5.

[Superzchen]

Generally, the DNA sequence and its complementary sequence are considered as two different sequences by the model. To solve this problem, I think you can incorporate both the DNA sequence and its complementary sequence as two separate samples in your training dataset. However, this still does not guarantee that two sequences will have the same score.