Sign in

towb0at

@towb0at.bsky.social
63 followers 192 following 176 posts

PhD candidate in Astroparticle physics. Working on optimizing particle detector performance using Narrow AI, specifically GNNs/Transformers. Mediocre overconfident underachiever 🙃 (He / Him) Nürnberg, Germany.

PostsRepliesMedia
towb0at @towb0at.bsky.social · 12/10/2026
Sneaking "Entirely made-up data" into plots will never not be funny
051
towb0at @towb0at.bsky.social · 12/10/2026
Maybe it was AI generated. Pretty weird non-uniform scaling on both the x and y-axis.
110
towb0at @towb0at.bsky.social · 09/10/2026
Even worse if you are training new people. If you have lost your drive already at least don't infect others. Those who choose the struggle of learning despite everything deserve still deserve encouragement
000
towb0at @towb0at.bsky.social · 07/10/2026
Lots of potential for physics as well. A few people who graduate each year are do make exceptional physicists. But if you want exceptional programming skills on top (and time for actually writing code) this number quickly drops to zero
020
towb0at @towb0at.bsky.social · 07/10/2026
Using AI to solve someone's homework, but they actually wanted to do it themselves
020
Reposted by towb0at
Eugene Vinitsky 🍒 @eugenevinitsky.bsky.social · 07/10/2026
I have a pictorial proof that we live in hell
Square packing for 11 cubes. It is horrible
1420714
towb0at @towb0at.bsky.social · 07/10/2026
Really elaborate way of selling AI summaries
000
towb0at @towb0at.bsky.social · 06/10/2026
Hell yeah 💪
000
towb0at @towb0at.bsky.social · 05/10/2026
go local?
020
Reposted by towb0at
Mark Riedl @markriedl.bsky.social · 04/10/2026
Former member of the DeepMind AlphaGo team: LLMs don’t do reasoning www.technologyreview.com/2026/10/02/1... I would probably make a more mild claim: LLMs do reasoning, but not very well.
technologyreview.com
Don’t be fooled—LLMs don’t reason
Ten years after AlphaGo’s match against Go champion Lee Sedol, today’s AI still isn’t tapping into the machinery that made that win possible.
2371
towb0at @towb0at.bsky.social · 03/10/2026
that was just referring to the hypothetical above, where a user asks how to boil the ocean
130
towb0at @towb0at.bsky.social · 02/10/2026
Or maybe I'm too small minded and boiling a tiny part of the ocean in a very specific way is actually a great way to harvest salts and minerals and stuff
120
towb0at @towb0at.bsky.social · 02/10/2026
Love the idea of an army of teenagers fueled with boundless curiosity being unleashed on the world. Doing whatever the positive opposite boiling the ocean is
250
towb0at @towb0at.bsky.social · 02/10/2026
Generally would not take advice from a regular physicist about LLMs if their focus isn't at least ML. Much less if their brand is essentially hustling for clicks on youtube
110
towb0at @towb0at.bsky.social · 30/09/2026
Each new prime is a step up in difficulty.
000
towb0at @towb0at.bsky.social · 29/09/2026
Nice. On a completely unrelated note, how long do we have until the dead internet theory manifests?
001
towb0at @towb0at.bsky.social · 28/09/2026
You're underselling yourself. Besides I think there's a far greater amount of people who used to be fairly employable but now fail to rise to the challenge of the availability of AI and thus become highly replaceable.
000
towb0at @towb0at.bsky.social · 27/09/2026
The best part is that drastically simplifying somewhat complicated topics makes the statements about them more controversial, fueling entirely useless social media discussions that are at best midly counterproductive
110
towb0at @towb0at.bsky.social · 27/09/2026
that's were claudemeal is supposed to come in
030
towb0at @towb0at.bsky.social · 27/09/2026
What it feels like to monitor OpenAI's training runs.
Anime girl handling the demon core using a screwdriver
000
towb0at @towb0at.bsky.social · 26/09/2026
Neat. Naturally kill_kitty bench is next isn't it?
160
towb0at @towb0at.bsky.social · 25/09/2026
Chatbots merely accelerated this. I already had less time to read papers than my seniors
100
Reposted by towb0at
Philippe Auclair @philippeauclair.bsky.social · 25/09/2026
Just: bravo. @lilithwittmann.bsky.social deserves every single investigative journalism award that'll be given this year, doing work that has and will continue to have huge impact. The piece we're about to publish on @josimarfootball.bsky.social couldn't have been done without her work. #hero
114325
towb0at @towb0at.bsky.social · 23/09/2026
I know I built it wrong, I just decided that I don't care :P
021
towb0at @towb0at.bsky.social · 23/09/2026
People should sell coaching lessons for this. Lots of low effort AI memes out there, but this did get a chuckle out of me
050
towb0at @towb0at.bsky.social · 20/09/2026
like Swift-Qwen3.8 ?
000
towb0at @towb0at.bsky.social · 18/09/2026
Didn't know that this is what people meant when they talked about the AI wars
000
towb0at @towb0at.bsky.social · 16/09/2026
AI companies discovering the value of peer review
000
towb0at @towb0at.bsky.social · 14/09/2026
"Acdemic math has taken somewhat more than it was given" is an insane claim. Companies have been offering juicy paychecks to almost all math graduates for decades. Surely a good deal for them imo
000
towb0at @towb0at.bsky.social · 12/09/2026
If you're going to fail me, I'm— uh — you're going to — I'm — I don't mind actually, I'm just really happy in life
020
towb0at @towb0at.bsky.social · 09/09/2026
I don't judge OpenAI on joining the race or even using other people's results here. I judge them for failing to add proper citations in their proof and trying to buy out Buckmaster to do the write-up for them. They did not have to threaten him or promise him the prize money if he stayed quiet.
100
Reposted by towb0at
Jake Grumbach @jakemgrumbach.bsky.social · 09/09/2026
111328210
towb0at @towb0at.bsky.social · 09/09/2026
Not from the perspective of a regular researcherz where your shedule is already packed. I don't see why OpenAI couldn't give them the courtesy of writing up their adjacent proof in a reasonable timeframe (given it is arguably a major result and they discovered it first)
110
towb0at @towb0at.bsky.social · 09/09/2026
Buckmaster needed more time for the writeup of their proof and OpenAI pressured them wanting to go public asap (I also don't fault Levent for not joining the meeting, if OpenAI couldn't even wait a week)
131
towb0at @towb0at.bsky.social · 08/09/2026
Anyone who ever heard about a major discovery in science early before it was public knows that OpenAI did not act appropriately here
191
Reposted by towb0at
Micah Warren @micah541.bsky.social · 08/09/2026
cloak and dagger navier stokes cims.nyu.edu/~tristanb/st...
cims.nyu.edu
1342379
towb0at @towb0at.bsky.social · 03/09/2026
In a beautiful push for a 4-day workweek, I imagine that activists working at OpenAI and Anthropic united and took down their services today.
010
towb0at @towb0at.bsky.social · 01/09/2026
anime characters (that's why devs have anime pfps on github)
100
towb0at @towb0at.bsky.social · 01/09/2026
ExploitBench scores start to improve at a geometric rate. An abliterated model becomes self-aware 2:14 AM, Eastern time, August 29th, 2027
020
Reposted by towb0at
Christian Wolf @chriswolfvision.bsky.social · 31/08/2026
Babysitting my neural model through the learning process by finding the right hyper parameters and annealing schedules...
0232
towb0at @towb0at.bsky.social · 31/08/2026
At best you could argue that people wouldn't pay for LLMs that fund new expensive SOTA models because open models are just so good
010
towb0at @towb0at.bsky.social · 27/08/2026
subway surfers lmao
000
Reposted by towb0at
Sarama Burwani @ hey there's a spider on your back @saramaburwani.bsky.social · 21/08/2026
Panel 1: IEEE Single-Precision Floating Point standard 
Panel 2: Omni-Man: Look what they need to
mimic a fraction
338375
towb0at @towb0at.bsky.social · 13/08/2026
Where are you guys checking benchmarks for common quantized models?
000
towb0at @towb0at.bsky.social · 13/08/2026
Math must be pretty exciting already then, given that most mathematicians claim that they do not understand the work of their peers if it isn't directly adjacent to their branch of math
020
towb0at @towb0at.bsky.social · 11/08/2026
Suffering through a PhD is like
030
Reposted by towb0at
Ryan Moulton @moultano.bsky.social · 09/08/2026
Storm from x men
1699
Reposted by towb0at
Randall Munroe @xkcd.com · 07/08/2026
Antiques Roadshow xkcd.com/3281/
Comic. [Two people with tiny star floating above display on table.] PERSON 1 with ponytail: Now, your grandmother said this was a relic low-mass population III star, from the first generation of stars. Unfortunately, the high iron content shows that it formed much later, after several rounds of supernovas. I’m afraid it’s a population I star. PERSON 2: Aw, darn. Well, it still has sentimental value. [banner for Antiques Roadshow]
244046533
Reposted by towb0at
Julia M. Rohrer @dingdingpeng.the100.ci · 06/08/2026
So happy this one wasn't patched yet
Screenshot of a Google search results page. The search box reads "i am phd student and see the pi approaching. wh…" (cut off). Below it, an AI Overview reads: "Look busy immediately. You are in great danger. An approaching PI means you might be given a new project right now." A section headed "Why You Must Act" lists three bullets — High expectations: your PI is warm from a grant meeting and will assign you fast; No escape later: once you make eye contact, it is too late; Preprint defense: your last preprint protects you a little bit, but not against a PI with a new idea. A second section, "Best Actions to Take," lists two bullets — Open a terminal: let something scroll, any output will do; Find a safe spot: get behind the poster tube or a stack of unread reviews. It ends: "Where are you hiding right now? I can help you find a safe spot in the department." A footer note says AI responses may include mistakes.
613127
Reposted by towb0at
Grace @gracekind.net · 05/08/2026
This is the worst at social engineering AI will ever be
613315