Zachary Charles (@MatharyCharles) header
Zachary Charles (@MatharyCharles) profile picture

distributed machine learning @ google | sometimes mathematician

Seattle
Joined September 2012
Photo by Zachary Charles (@MatharyCharles)Photo by Zachary Charles (@MatharyCharles)Photo by Zachary Charles (@MatharyCharles)Photo by Zachary Charles (@MatharyCharles)Photo by Zachary Charles (@MatharyCharles)Photo by Zachary Charles (@MatharyCharles)Photo by Zachary Charles (@MatharyCharles)Photo by Zachary Charles (@MatharyCharles)Photo by Zachary Charles (@MatharyCharles)Photo by Zachary Charles (@MatharyCharles)
Zachary Charles retweeted
Math would be a lot more fun right now if the models were better at exposition!
33
13
7
683
23,863
oof
Replying to @ElliotGlazer
Followup: Fails an Astra audit. Come on what were these guys doing the past few months🤦‍♂️
1
472
Zachary Charles retweeted
Btw Alessandro Masullo proposed a proof of GRH several months ago that has been under review by Brian Conrad "... and it is so far looking good." math216.wordpress.com/2026/1…
47
142
69
1,273
262,298
Zachary Charles retweeted
so we have @__alpoge__ 's work with claude about the Riemann hypothesis that just came out (see alpo.ge/remark.pdf) same day, there's also an announcement about a "all human, with no AI" proof of the same problem that I think it's important to know about (see math216.wordpress.com/2026/1…)
4
6
101
10,179
There could not be a more apt metaphor for the central tension in the math community these days than seeing this on a wordpress site of all things
1
156
Huge day for the guy I knew in grad school who claimed that all math was trivial, we just weren't working hard enough
11
367
wtf I had things to do today
Well, looks like @__alpoge__ & @claudeai are claiming a proof of the Riemann Hypothesis alpo.ge/remark.pdf
1
1
11
433
Quick LLM alchemy question: what learning rate schedule leads to this curve?
2
263
Chris is cracked - doing this right took him to the forefront of JAX & XLA, and he did it in a single summer. Very excited to see this blog post finally go out!
Finally seeing my summer internship project at Google go live as a blog post! 🚀 We implemented Streaming DiLoCo natively in MaxText for distributed LLM training across low-bandwidth, multi-metro networks. Details below 👇 developers.googleblog.com/di…
7
1
23
1,164
Zachary Charles retweeted
Finally seeing my summer internship project at Google go live as a blog post! 🚀 We implemented Streaming DiLoCo natively in MaxText for distributed LLM training across low-bandwidth, multi-metro networks. Details below 👇 developers.googleblog.com/di…
3
9
2
47
6,005
Zachary Charles retweeted
also can’t discredit that it’s “in vogue” to care about outages and cybersecurity incidents, so of course any outage that happens is going to get a lot of headlines, even if last year the same outage would’ve just made it to the org’s Status page and tech blogs
3
26
4
1,057
55,972
What's wrong with getting a little excited about all the good work you're doing?
Unhinged reasoning from Gemini 4 Carbon Why does it write like this? Genuinely curious
8
365
Zachary Charles retweeted
I've tweeted before about how architecture improvements usually improve the constant factor in loss scaling laws, keeping the same exponent. The scaling ladders in the Ripple-65B release from @RicursiveAI are a nice illustration of this at scale! 🧵
5
10
88
5,665
Zachary Charles retweeted
this is cool but it's extremely weird framing imo to call the stacks project "critical infra," that feels like it's importing intuitions from engineering that just don't apply here. if you deleted the stacks project tomorrow it's not like planes would fall out of the sky
Most think vibecoding at scale is slop. Here’s a counterexample. Today we release a 16M-line formalization of the entire Stacks Project, critical infra underlying modern math, with optimized code quality and its highly interconnected structure preserved. A mathematical city 🌆
12
3
1
144
9,019
Zachary Charles retweeted
Hi! this is Rohan Anil, I develop training algorithms among other things - my work trains various frontier models. At CoreAuto, I was fascinated by the problem of training on compute islands, eventually in space training. I hope you all enjoy this write up!🪐🛰️ coreauto.com/hitchhikers-gui…
66
137
39
1,581
193,743
Zachary Charles retweeted
For today's Sunday Review, we're reconnecting with a 2004 cult-classic video game whose postmodernist soundtrack doubles as a treatise on the meaning of life pitchfork.com/reviews/albums…
11
264
96
1,325
211,736
Zachary Charles retweeted
Tinker’s Dimuon is Muon with delayed diagonal conditioning, Polar Express and a rank-dependent update norm. I recovered its update rule through the API, borrowing @mrtnm’s approach to StableReinforce. Here’s how 🧵
Dimuon is a new optimizer we built for LoRA fine-tuning. In our SFT tests, it often reaches lower loss in fewer steps than AdamW. We're still experimenting with Dimuon; you can Tinker with it today and share your feedback! See our optimizers guide at tinker-docs.thinkingmachines…
14
31
2
355
20,759
I agree with this and adopted this stance unilaterally on a recent review. I encourage you to adopt it in your reviewing — it’s reasonable to write reviews that say “this is dense unreadable slop, my review isn’t detailed and I recommend rejection solely on that basis.”
Some thoughts on reviewing: spylab.ai/blog/reviewing/ First for @satml_conf where we trialed an early reject phase which cut submissions by ~40%. Then more generally about how not everything deserves peer-review, and why we should reward writing things that people actually read.
9
13
6
184
32,802
Now's actually the best time to submit great work to TMLR - good work shines through!
What is TMLR doing under a 300% increase in submissions? - Suspensions of >150 authors - Publicly naming major violations - Submission quotas distinguishing slop patterns - Desk rejections 6% → 53%; focus on clarity - AI reviews evaluating soundness blog.tmlr.org/2026/adapting-…
6
557
Zachary Charles retweeted
People really have no idea what string theory is, most have never talked to a string theorist nor bother to read up on it. Its public perception is driven by remarkably confident statements constructing a bogeyman that has nothing to do with the actual research
String theory is what happens when an intelligent system is starved of sensory input and starts hallucinating. Intelligence needs grounding. For this reason, intelligence is practically never the bottleneck to solving impactful problems. The bottleneck is iterated contact with reality.
116
59
11
902
54,769