@laxsill@social.spejset.org2026-08-06 19:11:11
ok gimp kan mycket men kan det köra doom? schack matt linux nördar https://feber.se/pc/kan-ms-paint-kora-doom/494763/
@arXiv_csAR_bot@mastoxiv.page2026-08-14 07:30:56
GateTruth: Auditing the Rigor of RTL Design Benchmarks via Mutation Testing
Meet Bhadra
https://arxiv.org/abs/2608.12635 https://arxiv.org/pdf/2608.12635 https://arxiv.org/html/2608.12635
arXiv:2608.12635v1 Announce Type: new
Abstract: Benchmarks for evaluating large language models on register-transfer-level (RTL) hardware design have proliferated rapidly, yet none reports having applied mutation testing, an established hardware-verification technique for quantifying testbench quality, to ask whether its own testbenches are trustworthy. A testbench that never fails is not evidence of a correct design; it may simply never stimulate the logic that is actually broken. We introduce GateTruth, a mutation-testing engine and methodology for auditing RTL benchmark testbench rigor: inject a deterministic, seeded set of semantic mutants into a reference design and measure what fraction the testbench catches.
We validate the methodology against our own 68-task, dual-track reference suite -- 60 specification-to-RTL generation tasks and 8 agentic-repair tasks, scored through a pinned, deterministic synthesis-to-timing flow with correctness enforced as a strict gate -- certifying that 46 of 60 Track A testbenches kill at least 95% of injected mutants under sequential, reproducible execution; we disclose why the other 14 do not, including a Goodhart effect on testbenches revised to pass this gate. We then point the same engine, unmodified, at RTLLM v2.0, a widely adopted external benchmark: of 46 auditable designs, 72% fall below the 95% floor our own suite is held to, and three score 0% outright.
A comparable audit of NVIDIA's CVDP benchmark is structurally impossible: its public release withholds reference solutions, removing the golden RTL mutation testing requires. Auditing our own instrument also surfaced a second finding: an initially uniform 4096-token output cap silently truncated three of seven evaluated models, and re-running at 16,384 tokens moved one model from fifth place to first. We argue mutation-kill certification should become a standard reporting requirement for RTL-generation benchmarks generally.
toXiv_bot_toot
@dnddeutsch@pnpde.social2026-08-22 13:58:16
Bei der Verleihung des Deutschen Kleinkunstpreises spielt Tobias Mann ein KI generiertes Lied um Preisträger Dr. Pop anzukündigen. Keine Pointe.
(ab ca. Minute 25)
https://www.zdf.de/play/shows/deutscher-klei…
@mlawton@mstdn.social2026-07-21 03:41:05
The air handler in the attic suffered a blockage in the drainage pipe. Water filled the pan and there is, apparently, no float kill-switch. And so, as water does, it sought the lowest point. Eventually that was the ceiling. 😱
A trip to the hardware store to by a wet vac and then a second trip to buy the foam filter required for wet vacuuming (in a frothy rage because why would you sell a wet vac without the ability to, I dunno, wet vacuum out of the box?!?)
Not a fun evening.
@hex@kolektiva.social2026-09-28 10:41:48
It occurred to me that "AI" is following the Walmart model. If they could keep token prices suppressed long enough to kill everyone able to use their brain, then they can recover the losses by being the only provider of those skills.
It looks to me like a move to swap LLMs for oil as the main tool of control, since everyone knows oil can't go on forever and there's no other choke point like it. They have to create a choke point or risk losing control of everything.
@thomasfuchs@hachyderm.io2026-08-02 03:24:17
I don’t buy the “AI will kill all jobs” argument, because you can (thoroughly and with high quality) automate only mechanical, repetitive and non-creative work.
For example you can’t automate fields of knowledge (e.g. mathematics), it doesn’t matter if it’s LLMs or some other yet to be invented machine doing it.
If we humans don’t know the knowledge (and understand it), it might as well not exist.
Some might claim that “AI-powered fully automated luxury abundance” is the future, and we don’t need to know anything.
First of all, “AI” doesn’t exist. It’s science fiction. LLMs aren’t intelligent and can’t think.
More importantly though, many people want to know and understand things.
What’s the point of a world that’s synthetically generated based on what past humans did and where no one knows anything or learns anything anymore?
@tomkalei@machteburch.social2026-07-24 14:30:30
It's not even clear what these people even want from us? Academic recognition? Co-Authorship on some follow-up paper? A cookie?
I think behind all this is a gross misunderstanding of how mathematics works and what the day-to-day conjectures in our papers mean.
They are not huge prizes like Jacobian conjecture or similar but they would serve as a little prize for the humans (often students!) who do point 2.
What do we do now? Keep our conjectures for ourselves? 2/2
https://thomas-kahle.de/blog/2026/just-42-tells-us-nothing/
@BBC6MusicBot@mastodonapp.uk2026-10-01 09:43:46
🇺🇦 #NowPlaying on #BBC6Music's #LaurenLaverne
Little Simz:
🎵 Point and Kill (feat. Obongjayar)
#LittleSimz
https://aguava1.bandcamp.com/track/little-simz-point-and-kill-feat-obongjayar-aguava-edit
https://open.spotify.com/track/52y4KhkcAbYcogFg2u7UVP
@tiotasram@kolektiva.social2026-08-25 22:20:42
I Can See the Babies
---
Whenever I see a person, my mind automatically conjures up an image of them as a baby. It's not really voluntary, nor is it the same as other mental imagery. More like I'm hallucinating a visual echo of their past self, floating in my vision near where I see the person. As if they were haunted and I could see the ghost.
It's all in my head, I'm pretty sure, and the babies aren't necessarily accurate. I checked once with a friend, and the baby I saw had curly hair even though all their baby pictures were bald. After seeing those pictures, my baby image of them changed, too, to match the pictures. So that's why... That's why I'm pretty sure...
But the babies are real, you know? I mean, my images of them might not be right, but all the people I see *were* babies. It's not like one of those stupid prophecy loopholes like "everyone was born of woman", you see, it's a true universal, no exceptions. And these hallucinations won't let me forget about that. Ever.
I tried to play a videogame once, one of those FPS games. My friends were all playing and they needed one more for their group. They didn't know about... this. I told them I didn't think I could do it, but I didn't want to tell them why, so they badgered me into it. I thought maybe I could handle it, I mean the characters were basically cartoons, right? But... well, I think I dissociated a bit and then threw up. Even if they were cartoons, all those babies, shooting and killing and dying, again and again. It's... it's not like the babies got injured, even, or played out exactly what the characters were doing. The babies just kinda floated around doing baby things, and they sort of faded out when their character died. No, what's overwhelming is the knowledge, the constant reminder, of everyone's former innocence.
The worst part, you know, the most horrible thing, is not just the innocence, it's the gap. Like, if you were to force me to watch a torture scene, I would hate seeing the victim, but it's the torturer that scares me. Everybody knows that horrible things happen to innocent people. We're all forced to think about that from time to time. But the thing almost nobody thinks about as much as me, is how innocent babies *become* monsters. How there's a hurricane of hurt and fear and apathy and indoctrination and hatred and prejudice and complicity out there that sucks in babies and twists them into racists and abusers and "business leaders" and soldiers and all sorts of people who hurt and kill others. The worst are the ones who take pleasure in that hurting, because the babies for sure don't. Like, how did that happen? And most people don't think about it, because they just label that person as "bad" and don't think about it any deeper. But you can't do that when you can see the baby right there. The baby isn't bad. The person might be, I'm not trying to say everyone's an angel or people don't deserve to reap what they sow. But the baby wasn't! Something between then and now went wrong, horribly wrong, and now you're stuck with a person that hurts others, and you need a drastic intervention just to stop them, let alone prevent them from doing that stuff again and change their future behavior. But you can *see*, or at least, I can see, that there was always some earlier point, some missed opportunity, where some much easier intervention could have headed the whole thing off. None of those babies are destined to turn into monsters, and that's the gap. The gap between the little baby that was and the monster in front of you on the screen or in real life.
Anyways that's it. That's my statement. I'm not a pacifist or anything, but I think you can see why I can't do it, I just can't. You can lock me up or punish me or whatever, but there's just no way you're gonna put a gun in my hands and give me an order and I'll just follow it, because I can see the babies.
#Microfiction