What do you mean "the hot remains?"

Might be a multiwire branch circuit (look carefully at the wires coming off breaker A and B...do they go into the same cable?).

I am using a non contact voltage tester. When both circuits are energized the tested emits a sound as expected. When i turn breaker A off the light turns off however the non contact voltage tester still emits a sound. Only when i turn off breaker B nothing is detected by the voltage tester. I will break out the DMM later today and also take a look at the box.

These non-contact testers (and even some of the contact testers) yield false positives a lot. When you have a long disconnected wire it will couple enough to an adjacent wire to give a reading. Circuit B just likely is the live one that it was coupling to (They were close enough together for part of the run).

Note that most digital meters are sufficiently high impedence that they can be fooled as well.

Thanks for the info after measuring voltage it was 0v. So i will take the non contact testers for what they are worth.

I have some analog DMMs. It is just easier to grab a solenoid tester. I like my Triplett Jolt-Bug, but I used to employ a light bulb. Both methods are faster than any meter, yes-no. There is little judgment or thought (no dial or display to wait for response and to mindfully integrate).