Results on Virtual distillation circuits for error mitigation

Hello Pennylane help

I have a problem with the result of my calculation about VIrtual Distilation mitigation that I read from this paper, in february this code produces the first graph but now when I run it the result is on the second graph where the mitigation result can’t be better than noisy. Can you help me to fix this code, I am very grateful for your help.

you can see this code here

first:

second:

Hi @Dwi_Cahyo_Mariyanto ,

Can you please clarify what you mean? The only difference that I see is a very tiny improvement in the first graph with the mitigated models vs the noisy one. But the difference looks so small that it could be a matter of randomness in the noise. If the noise is now a tiny bit smaller, that could explain why the second graph no longer shows that tiny gap.

Since February we have only released one major version of PennyLane, v0.41, which is the current version. This didn’t add any changes to QNG. It may have introduced changes in the underlying components but this is unlikely to be the cause for the difference you’re seeing. You can take a look at the release notes to double-check, but I didn’t find anything.

I hope this helps.