Timo Hillmann
timohillmann.bsky.social
Timo Hillmann
@timohillmann.bsky.social
PhD student @ Chalmers University of Technology
Thanks for this collaboration @lucasberent.bsky.social, Armanda Quintavalle, @jenseisert.bsky.social, Robert Wille, and @qec.codes (Joschka Roffe)
September 3, 2025 at 3:47 PM
LSD can get its soft information from a pre-processor like
BP or the device's noise model directly, see also VibeLSD. arxiv.org/abs/2508.15743
It matches the performance of the best decoders out there, like OSD-O, but with a reduced runtime complexity.
Colour Codes Reach Surface Code Performance using Vibe Decoding
Two-dimensional quantum colour codes hold significant promise for quantum error correction, offering advantages such as planar connectivity and low overhead logical gates. Despite their theoretical ap...
arxiv.org
September 3, 2025 at 3:47 PM
All of this is made possible through a novel parallel matrix factorization strategy, which we call on the-fly elimination, to identify, validate, and solve local decoding regions on the decoding graph.
September 3, 2025 at 3:47 PM
I think I’m missing :)
June 30, 2025 at 2:09 PM
Since even for surface codes Relay-BP (Fig. 2) competes with matching-based decoders, it would be interesting to see whether Relay-BP outperforms known color code decoders, relevant, for example, to the performance of magic state cultivation.
June 3, 2025 at 8:28 AM
As Gallager already noted, BP is often more art than science, justified “only by the fact that it works.” But it does work in many cases, and I think it’s worth exploring how far we can push it with careful heuristics and structured approaches.
June 3, 2025 at 8:28 AM
Another thing I noticed: there has been only limited effort to combine different strategies to mitigate trapping sets. I had actually written this down as a future project idea but didn't get to it, so I’m glad to see that the IBM Quantum team has already shown that this direction is viable.
June 3, 2025 at 8:28 AM
While writing my thesis, I revisited a lot of the BP literature and dug into the basics. One thing that stood out is that heuristics for trapping sets are rarely tested under circuit-level noise, even though decoding in that setting is significantly different from code capacity.
June 3, 2025 at 8:28 AM
Reposted by Timo Hillmann
There is now an official announcement: The TIB has archived the current status of the arXiv:

blog.tib.eu/2025/05/14/p...
Protecting Science: TIB builds Dark Archive for arXiv - TIB-Blog
Research and science are international; it is not for nothing that we speak of international specialist communities. Although a service such as arXiv is operated by an institution based in the USA, na...
blog.tib.eu
May 14, 2025 at 2:50 PM
Did the text change significantly? If I recall correctly the previous ones then they read very similar. Only the word „exceptional“ didn’t appear every time if my mind is not lying to me
May 7, 2025 at 8:36 PM
If now for subsystem codes we are measuring non-commuting gauge operators and combining them to get stabilizers, there is a way to look at the problem in the same way as for the ordering of operations in ordinary syndrome extraction circuits. Probably with even fewer constraints though. 2/2
April 7, 2025 at 4:17 PM
I wonder if with hindsight we are just coming back to realize that the order of operations in the syndrome extraction circuit matters for the effective fault distance of ordinary codes. There we all agree that the order of operations matters. 1/2
April 7, 2025 at 4:17 PM
Section 3 of the Extractor paper is a very good overview I find as a non-expert. arxiv.org/abs/2503.10390
Extractors: QLDPC Architectures for Efficient Pauli-Based Computation
In pursuit of large-scale fault-tolerant quantum computation, quantum low-density parity-check (LPDC) codes have been established as promising candidates for low-overhead memory when compared to conventional approaches based on surface codes. Performing fault-tolerant logical computation on QLDPC memory, however, has been a long standing challenge in theory and in practice. In this work, we propose a new primitive, which we call an $\textit{extractor system}$, that can augment any QLDPC memory into a computational block well-suited for Pauli-based computation. In particular, any logical Pauli operator supported on the memory can be fault-tolerantly measured in one logical cycle, consisting of $O(d)$ physical syndrome measurement cycles, without rearranging qubit connectivity. We further propose a fixed-connectivity, LDPC architecture built by connecting many extractor-augmented computational (EAC) blocks with bridge systems. When combined with any user-defined source of high fidelity $|T\rangle$ states, our architecture can implement universal quantum circuits via parallel logical measurements, such that all single-block Clifford gates are compiled away. The size of an extractor on an $n$ qubit code is $\tilde{O}(n)$, where the precise overhead has immense room for practical optimizations.
arxiv.org
April 7, 2025 at 4:09 PM