which in turn are a variant of classical belief propagation decoding.

Thanks to @joerenes.bsky.social for the really enjoyable collaboration!

3/3

Our algorithm can be used to optimally(!) decode any classical linear code with an efficient trellis representation that is transmitted over a binary-input classical-quantum pure-state channel. As such, it can be considered a quantum analogue of the classical Viterbi/BCJR algorithms, ...

2/3

Really excited about our new paper on quantum belief propagation (aka "belief propagation with quantum messages")!
arxiv.org/abs/2509.19441
Our new quantum algorithm solves a structured state discrimination problem which is central for quantum algorithm based on Regev's reduction (like DQI).

1/3

Finally, I'm super excited to announce that I will soon start a postdoc at Inria Paris in the COMSIQ group, funded by a SNF Postdoc.Mobility grant!
@snf-fns-ch.bsky.social

(4/4)

Circuit knitting protocols have gained some attention in recent years, because some people hope they could help us to do something useful with near term quantum devices. See for example www.nature.com/articles/s41... (work not by me) which experimentally implemented some of our protocols.

(3/4)

These techniques allow you to do cool things, like simulating a large quantum computer with smaller ones (usually called "circuit knitting" or "circuit cutting"), by dividing the circuit into smaller chunks. The simulation overhead grows exponentially with the amount of non-local interactions.
(2/3)

Hey all! Just uploaded my PhD thesis on the arXiv, which I recently defended in January. Feel free to check it out! 🤗

arxiv.org/abs/2503.21773

It's about quasiprobabilistic algorithms to simulate quantum circuits - not just on classical computers, but also on restricted quantum computers.

(1/4)

Congratulations! 😊