Alex Meiburg Profile
Alex Meiburg

@Timeroot

Followers
160
Following
1K
Media
156
Statuses
1K

Quantum info, Magic the Gathering jokes, and whatever else. | Postdoc @ UWaterloo IQC, Perimeter Institute.

Joined July 2009
Don't wanna be here? Send us removal request.
@Timeroot
Alex Meiburg
6 days
Damning
@alexcdot
Alex Cui
6 days
There are a LOT more papers that need to be desk rejected at ICLR. Somehow, this hallucination wasn't caught. So, I went on a crazy rabbit hole and found 50 more (many are just as funny). We're approaching crisis levels Paper titles included below 👇 https://t.co/PeTjJAbkrH
1
0
2
@Timeroot
Alex Meiburg
12 days
Stochastic parrots, amirite? Just ... um ... regurgitating what's already out there ... right? 🤔🤔🤔
@vladtenev
Vlad Tenev
12 days
We are on the cusp of a profound change in the field of mathematics. Vibe proving is here. Aristotle from @HarmonicMath just proved Erdos Problem #124 in @leanprover, all by itself. This problem has been open for nearly 30 years since conjectured in the paper “Complete sequences
0
0
3
@vladtenev
Vlad Tenev
12 days
We are on the cusp of a profound change in the field of mathematics. Vibe proving is here. Aristotle from @HarmonicMath just proved Erdos Problem #124 in @leanprover, all by itself. This problem has been open for nearly 30 years since conjectured in the paper “Complete sequences
262
658
5K
@nasqret
Bartosz Naskręcki
18 days
I have personally crossed a line, and I am a bit in awe. This is my first fully automated, LLM-generated and auto-formalized proof of a new mathematical theorem. Let me set up the problem: we have three rotating circles with six positions each, all three intersecting in a total
53
174
1K
@Timeroot
Alex Meiburg
1 month
Pet peeve: browser tabs that capture standard keyboard shortcuts. Case in point: Juypyter notebook. Ctrl+Shift+T: reopen last tab. Doesn't work now. Ctrl+Shift+S: screenshot (on FF). Doesn't work. A litany of others. I'm sure there's settings to fix this, but it's aggressive.
0
0
2
@Timeroot
Alex Meiburg
1 month
I thought the irony was lethal when Monopoly became a big seller. But no, it can always get worse! https://t.co/k5q8yoVkhX
polygon.com
In search of free food and rare Monopoly game pieces, McDonalds customers are dumpster diving and visiting multiple stores
0
0
0
@Timeroot
Alex Meiburg
1 month
> Lizzie Maggie: hates how capitalism accumulates wealth and hurts people > Makes board game about it > "Haha collecting money is fun!" > Government cuts SNAP benefits > Board game promo for free McDonalds > People play anti-capitalist capitalist board game so they don't starve
1
0
0
@Timeroot
Alex Meiburg
2 months
The section "Reducing tensor network costs via gate removal" from the supplementary material talks about one classical algorithm they could run that introduces extra error, but speeds it up. But they still get lower error than the quantum computer and it takes 5 min, not 3 years.
0
0
0
@Timeroot
Alex Meiburg
2 months
Can someone explain to me figure S38 from the supplemental material? Because it *really* looks like (also in the accompanying text) Google is saying that they can actually do it faster on a classical computer, getting better accuracy. "13000x" is only for exact calculation. What?
@GoogleQuantumAI
Google Quantum AI
2 months
Building quantum computers faces a core challenge: control qubits without losing info. Hear from Yu Chen, Director, Quantum Processor, on the “secret sauce” behind Willow. It let Quantum Echoes run 13,000x faster, a verifiable quantum advantage. Read → https://t.co/6mzhF78AQp
1
0
0
@Timeroot
Alex Meiburg
2 months
That's not once, but *twice* in 24 hours that someone announces that a Mathematician+ChatGPT make a notable discovery, and then AI formalizes it into Lean.
@llllvvuu
L
2 months
@ErnestRyu I inputted these pics into @HarmonicMath Aristotle and it produced a formally verified proof of the claim in Lean 4:
0
0
2
@Timeroot
Alex Meiburg
2 months
steps that we can trace that show the original idea (a literal thread) became the modern concept (a series of choices of deductive attack), so in the end we can follow the path. (4/4) (Are my jokes funny yet?)
0
0
0
@Timeroot
Alex Meiburg
2 months
it is in the end the same concept, because it can be traced to the same origin and has the same name. It is simply a concept which has had each constituent notion replaced. Confusing? Yes, but the literature about the term gives a continuous series of evolutionary (3/)
1
0
0
@Timeroot
Alex Meiburg
2 months
received from Ariadne to help him out of the maze. The original meaning has nothing in common with the modern logical concept, as each component of this metaphor about Theseus has been gradually replaced with a different and more abstract idea than a literal thread. **But** (2/)
1
0
0
@Timeroot
Alex Meiburg
2 months
In formal logic, "Ariadne's Thread" refers to a specific series of deductive steps that leads you from a beginning to a conclusion, telling you which way to proceed when there are multiple reasonable steps to take. The name is an allusion to the thread that Theseus... (1/)
1
0
0
@Timeroot
Alex Meiburg
2 months
You take 100% of the shots you miss.
0
0
0
@Timeroot
Alex Meiburg
2 months
@HarmonicMath across a number of domains, including our ongoing formalization of quantum information. I'm excited to get this paper in front of people - and we're even more excited to get the AI out to public access... soon! (/end)
0
0
1
@Timeroot
Alex Meiburg
2 months
@HarmonicMath this system is able to help with *lots* of other kinds of proofs. We give some examples in the paper, material that is *very* much not high school. In my own professional capacity as a mathematician, it's considerably accelerated my ability to write proofs, (2/)
1
0
1
@Timeroot
Alex Meiburg
2 months
New paper out! https://t.co/4eKUJ8fij0 This one's with @HarmonicMath. We've trained an AI that was able to crush the IMO, the International Math Olympiad, in an entirely rigorously checked way in Lean4. But even more importantly, we've shown that... (1/)
Tweet card summary image
arxiv.org
We introduce Aristotle, an AI system that combines formal verification with informal reasoning, achieving gold-medal-equivalent performance on the 2025 International Mathematical Olympiad...
1
1
8
@Timeroot
Alex Meiburg
3 months
Lean's Mathlib has a definition of the "positive part of a matrix". It is a sufficiently complicated definition that, by default, Lean fails to load it. You need to increase the max processing steps, otherwise it simply quits.
0
0
1
@Timeroot
Alex Meiburg
3 months
Log on to X for first time in a while A bunch of people, including Americans, are upset about something A guy got killed A guy who lived by and ruled with the motto "Death to America" Confusion I log off X
0
0
0