Lightnews — Scholar-powered news

anthony-leverrier.bsky.social @anthony-leverrier.bsky.social · 3d

Kitaev also

2

anthony-leverrier.bsky.social @anthony-leverrier.bsky.social · 8d

grandfather? he's not *that* old
;-)

1

Reposted by anthony-leverrier.bsky.social

John C. Baez @johncarlosbaez.bsky.social · 13d

There's no reachable position in chess with more than 218 moves for one player! Here's one with 218 moves for White.

This was *not* proved by checking all 8.7 × 10⁴⁵ reachable chess positions. Instead, someone known only to me as Tobs45 figured out a better proof:

lichess.org/@/Tobs40/blo...

A weird chess position where White has 9 queens, two rooks in the corners, etc, and black is checkmated with king in the corner, and only two pawns. From here:

https://lichess.org/@/Tobs40/blog/there-is-no-reachable-chess-position-with-more-than-218-moves/a5xdxeqs

3 3 23

anthony-leverrier.bsky.social @anthony-leverrier.bsky.social · 15d

this is apparently true!

en.wikipedia.org/wiki/Taiping...

1

Reposted by anthony-leverrier.bsky.social

Dulwich Quantum Computing @dulwichquantum.bsky.social · 17d

Stephen Colbert explains why the simplest way of finding the shortest path between 22 cities involves
1. flying to the Moon
2. mining some Helium-3
3. bringing it back to the Earth and pouring it in your quantum computer.

1 3 15

anthony-leverrier.bsky.social @anthony-leverrier.bsky.social · 17d

playing with chatGPT myself quite a bit, I see it make sentences like "It presents a drop-in, provably correct replacement for the contested Step 9" quite a bit, so I'm also a bit suspicious here.

2

anthony-leverrier.bsky.social @anthony-leverrier.bsky.social · 19d

This is interesting and surprising: some Clifford gates cannot be implemented arbitrarily well on arbitrarily good approximate GKP states with Gaussian operations!

Nice work by Brenner, Diaz and Koenig
arxiv.org/abs/2509.14658

Cat qubits (and generalizations) for the win, I guess
;-)

Composable logical gate error in approximate quantum error correction: reexamining gate implementations in Gottesman-Kitaev-Preskill codes

Quantifying the accuracy of logical gates is paramount in approximate error correction, where perfect implementations are often unachievable with the available set of physical operations. To this end,...

arxiv.org

1 14

anthony-leverrier.bsky.social @anthony-leverrier.bsky.social · 20d

thinking model with the 20$ version

anthony-leverrier.bsky.social @anthony-leverrier.bsky.social · 20d

So, not a breakthrough, but something that could have kept an intern busy for a while a couple of years ago. (And not something I would ever have attempted myself before AI tools were available.)
(4/4)

3

anthony-leverrier.bsky.social @anthony-leverrier.bsky.social · 20d

but it ended up working very nicely.
One gets quite decent parameters, better than what I thought.
@eczoo.bsky.social (3/4)

1 2

anthony-leverrier.bsky.social @anthony-leverrier.bsky.social · 20d

compute the code parameters. Nothing new, this is exactly the construction for the original paper but coding all of this can be rather painful.

With ChatGPT, i just had to explain the idea and ask for the scripts that compute the parameters. It took some time (about at day in total) to debug (2/4)

1

anthony-leverrier.bsky.social @anthony-leverrier.bsky.social · 20d

Can ChatGPT help with research? Maybe not yet for finding new results, but it can certainly speed up some tedious tasks.

An example about quantum Tanner codes #qLDPC
Working with qudits (d=5) lets you use nice local codes [4,2,3]_5. Then the idea is simply to enumerate small groups and (1/4)

7 4 15

anthony-leverrier.bsky.social @anthony-leverrier.bsky.social · 20d

sure, and I would agree with that point. But showing the PQ line converging to probability 1 after a few years is a way stronger claim.

2

anthony-leverrier.bsky.social @anthony-leverrier.bsky.social · 21d

I think this is a very misleading figure. The consensus is that RSA will be broken with proba 1 once a quantum computer is available, but I don't think anybody reasonably expects that all postquantum protocols will eventually get broken.

From arxiv.org/abs/2509.13405

2 1 11

anthony-leverrier.bsky.social @anthony-leverrier.bsky.social · 21d

right... so what do we do?

1

anthony-leverrier.bsky.social @anthony-leverrier.bsky.social · Aug 29

what was the conclusion of the discussion?

4

anthony-leverrier.bsky.social @anthony-leverrier.bsky.social · Aug 27

Félicitations Hanna !!

1 1

Reposted by anthony-leverrier.bsky.social

Error Correction Zoo @eczoo.bsky.social · Jul 18

We are now at 999 codes. What should be our 1000th? If we pick yours, we will feature you in a post about the code.

Comment, reply, message, and spread the word.

2 3 9

Reposted by anthony-leverrier.bsky.social

Robin Kothari @robinkothari.bsky.social · Aug 6

The QIP 2026 call for papers is out! QIP 2026 will be held in Riga, Latvia from January 24–30, 2026. See you there!
qip2026.lu.lv

15 27

Reposted by anthony-leverrier.bsky.social

Andy Matuschak @andymatuschak.org · Aug 5

In what senses is the quote true or useful? One thing I like about it is that it takes for granted that the listener (me!) can work on great ideas, make a great contribution. Most work doesn't even try. But it's worth trying. The sentiment makes me want to set my sights higher.

1 1 10

Reposted by anthony-leverrier.bsky.social

Henry Yuen @henryyuen.bsky.social · Aug 2

Come for the iconic papers and eye-wateringly beautiful textbooks, stay for the stories "from the trenches" (of which I hope John posts more of!). Keep writing, @johnwatrous.bsky.social !

John Watrous @johnwatrous.bsky.social · Aug 2

Bluehost seems to be holding my domain name hostage, I gather in an effort to sell me more products and services. Avoid them at all costs.

But see if I care. Hereafter you can find my web page at jhwatrous.github.io in case you're looking for it.

John Watrous @johnwatrous.bsky.social · Jul 23

I ditched my old website and created a new one on GitHub Pages:

johnwatrous.com

It's a great option for a static web page, particularly if you like to do your own CSS styling rather than flipping through pre-built templates that are never quite right.

2 7

Reposted by anthony-leverrier.bsky.social

Craig Gidney @craiggidney.bsky.social · Jul 30

When their preprint came out, it was overshadowed by Regev's paper. I found it disconcerting: space is the key near-term cost and Regev was paying space while this paper was saving space, but then Quanta and Scott Aaronson mentioned Regev but not Chevignard+. Glad they're getting recognition now.

1 1 6

Reposted by anthony-leverrier.bsky.social

Craig Gidney @craiggidney.bsky.social · Jul 30

Chevignard's QIP2025 talk on reducing the space cost of quantum factoring:

www.youtube.com/watch?v=R3yM...

And her co-author Schrottenloher's talk at the Simons Summer Cluster on Quantum Computing:

www.youtube.com/watch?v=0pnX...

Clémence Chevignard: "Reducing the Number of Qubits in Quantum Factoring" (QIP 2025)

YouTube video by QIP2025

www.youtube.com

1 1 19

Reposted by anthony-leverrier.bsky.social

Pixels | Le Monde @pixelsfr.bsky.social · Jul 21

Gilles Dowek, informaticien engagé et vulgarisateur, est mort

Passionné par la dimension éthique de sa discipline, le chercheur de l’Inria et professeur attaché à l’ENS Paris-Saclay, est décédé lundi 21 juillet, à l’âge de 58 ans.

www.lemonde.fr

20 24

Reposted by anthony-leverrier.bsky.social

Dulwich Quantum Computing @dulwichquantum.bsky.social · Jul 8

"Awesome Quantum Computing Experiments: Benchmarking Experimental Progress Towards Fault-Tolerant Quantum Computation"
arxiv.org/abs/2507.03678

A comprehensive database of notable quantum computing experiments, with emphasis on quantum error correction implementations:
github.com/francois-mar...

Awesome Quantum Computing Experiments: Benchmarking Experimental Progress Towards Fault-Tolerant Quantum Computation

Achieving fault-tolerant quantum computation (FTQC) demands simultaneous progress in physical qubit performance and quantum error correction (QEC). This work reviews and benchmarks experimental advanc...

arxiv.org

1 3 20