TLDR: An interested audiologist can find you out.
I also have tinnitus and there is a difference between hearing that a beep is played and hearing the beep. Hearing the beep means that you can clearly hear it and know for instance its pitch. Knowing that it is played means that you clearly perceive that the soundscape is different, you hear something repeat, but are not entirely clear on the matter. If you go from louder to softer, you’re more likely to understand those sound as the correct tones, than if you go from softer to louder: Very hard to discriminate the beep from the tinnitus for instance.
For me the difference between those two is typically 5dB. As the test calls for the threshold hearing level, basically where you can hear the beeps for 50% of the time, I would guess that the lower value is what is asked by the test.
However, I usually respond to the audiogram test with the clear answer: When I hear a beep clearly.
My audiologist did an extensive test with word recognition and REM and basically told me that the audiogram was about 3dB too high for optimal hearing.
So there you have it. Even with as little as 5dB difference between your minimum and maximum answers (The test itself claims only 10dB accuracy.) an interested audiologist can find you out.
So why would you want to change your audiogram? Well, for instance, when I was programming my previous aids with DSL (Desired Sound Level), I found that the words were not that clear as I needed them. Changing the audiogram from knowing to clearly hearing, solved the issue much better than just increasing the values by 5dB.