nate

I wrote an informal recap of the event on the UF blog: unitary.foundation/posts/2025_w...

A more technical summary is in the works as well, so be on the lookout for that next month :)

2025 and 2022 by the looks of it: qipconference.org/previousqips

🪦

Congrats @grageragarces.github.io!

I like toml but I'm too used to yaml and it's upsetting

On top of this, I can't unsubscribe from all newsletters at once without MAKING AN MLB.COM account... wtf

@mlb.com why, when unsubscribing to your 438925 separate email newsletters do I have to get verified to be human by cloudflare? Adding another step that makes it that much more inconvenient. Really annoying dark pattern.

damn i've walked by that place a few times and always thought it was sus. next time, if it's still going to be open i guess.

Pic of a math textbook where a proposition closes with “in particular, V is free as a k-module”

We should start saying things are free as a k-module more often

Modern is bomb I go there every time I’m in the city

WERQSHOP is a wrap! Glad to have had the opportunity to chair the event but damn that was a jump into the deep end in terms of event organizing.

Thanks to all who helped organize and to those who came :))) <3

See you soon!

Never see slides this good at academic conferences

cartoon of someone with awesome hair going real fast in a car where the driving wheel is the python logo and the rust crab logo is the engine.

Had a lot of fun presenting at SciPy today about noise in quantum computers. High-level talk about noise, and how python plays a role in speedy development. Slides are available here nates.place/static/scipy..., but this is my favorite one.

agreed, loglog is better.

anyone in #seattle tryna co-organize an html.energy day?

"we dive in this quantum pool

we work with it

we come out of the other end of the pool, and we're on dry land again. We're not talking quantum anymore and we've proven something that we didn't know before."

@merriam-webster.com dictionary claims that `unassign` isn't a verb...

Wiktionary (en.wiktionary.org/wiki/unassign) new best friend

Ayyy dats me. I guess I should be making some slides

I did my best to draw the true data which looks like this.

it's surprising to me that there isn't an easier way to get the full download history of your package than to use google bigquery

it's easy, but would've thought PyPI would have this all open somehow

I'm looking forward to WERQSHOP (werq.shop) for this very reason to have the opportunity to get to discuss what we as a community really need in this space over the next ~5 years to get EM into error corrected/fault tolerant machines

I've tried to help EM researchers contribute code to mitiq, but it doesn't align well with their "job to be done" which is usually producing ✨ new ✨ ideas.

Ideally mitiq would be the place where that happens, but we (@unitary.foundation) don't have the eng hours/$ (*shakes fist at feds*) to implement + benchmark all the techniques/optimizations.

As a field there are still a ton of unknowns here around what techniques are effective in different noise regimes, technique composability, and in particular there haven't been enough EM benchmarks to fully understand the tradeoffs of each technique.

The main goal would be to get error mitigation (EM) researchers to contribute new ideas to the package (with help from us), which we haven't fully realized, but we have managed to add new techniques and features over the past year regardless.

When we brainstormed OKRs for the year, I thought our internal goal of hitting 100k downloads this year was a bit too ambitious. We had 80k last year, and a decline in people contributing to the project, but we seem to be right on target!

hand drawn plot of the number of downloads of mitiq over time. started early 2020 and hit 100k in mid 2023, and halfway through 2025 have hit 250k. potential futures go off to the right with lots of question marks

mitiq hit 250k downloads! Even with all the hype around error correction it seems any other way to mitigate the impact of noise is still important for the time being.

The next release (0.46.0) will come out soon! Keep your 👀 peeled.

github.com/unitaryfound...

Nice talk yesterday from Mariia Mykhailova (recently moved to @psiquantum.bsky.social) on software best-practices in quantum to kick off the UW REU seminar.

I liked Maria’s point that there is an elevated need to understand runtime ahead of running programs because feedback loops are so long

@physreva.bsky.social booooooooooooo

massive L

Yeah, i tried that earlier, but needed it to do a full rotation. It does help me understand that particular region which is handy.

My feature request is to add a slider to this parameter that allows me to toggle the value of t manually. (Not expecting you to implement this, just thinking out loud)

You mean run it on my slower quantum computer?