Ahh good paper choice. So many of my former colleagues on this one 🤩
09.03.2026 17:53
👍 1
🔁 0
💬 0
📌 0
Ahh good paper choice. So many of my former colleagues on this one 🤩
While I recognize the importance of scientific writing and precise terminology it would be nice if we made papers more approachable for those of us who don’t read Ramanujan’s notebooks in our free time 💀
Haha yeah I know, this was more of a vent post than a real question. In this case it’s just too much math terminology that I’m unfamiliar with despite knowing the underlying concepts.
I will get through it. But I will also curse the authors the whole time 🙃
Am I allowed to tell an AC that I'm too stupid to review a paper?
Guys I never expected to go from watching this guy play minecraft as a kid (this is why I learned how to code btw) to watching him train LLMs. What the actual fuck: youtu.be/aV4j5pXLP-I?...
When I was TAing regularly I was always astonished at how bad some of the cheats were at covering up… why cheat so obviously? I have come for realize that it’s just laziness
If you can put in the work to cheat properly you could probably just do the work lol.
The last time I can remember a public official who was this effortlessly cool is Obama.
Made a dedicated account for photography. Best viewed on @flashes.blue :)
Anecdotally it also seems that activity on bsky comes in waves. People get all excited about leaving Twitter after "the latest nazi thing" they come to bsky, and then they leave here after a week or so. I have 3k followers here but still get more engagement on Twitter where I have 300.
I suspect that graduate students are still incentivized to "obtain a following" in order to promote their work, and I suspect they feel like Twitter is still the best place to do that, especially with a lot of Industry folk still there rather than here.
The Discord age verification discourse focuses on adults losing access to communities they have a right to access. That's important.
I'm also worried, though, about the kids. They seem to be an afterthought in these conversations. They have much to lose too. 🧵 www.theverge.com/tech/875309/...
Inside you there is an agent swarm
A piazza post where the students are asking for an autograder with unit tests and the TAs have responded by savagely roasting the students for asking, emphasizing that in the real world you must write your own unit tests.
@eugenevinitsky.bsky.social this reminds me of a legendary post made by my TA (and future collaborator) Kyle. Clearly you’re a lot nicer than my instructors ever were 😂
Yeah gradescope is great! I used it when I TAd classes at UMass :)
I would be deeply disappointed if Gavin is the nominee
I would still vote for him, and so should you
Have heard good things about CachyOS (dunno if it counts as weird enough tho). Otherwise why not try vanilla arch + a fun WM? I’ve been enjoying hyprland a lot recently. Tiling is amazing if u have a big monitor.
Unfortunately as it turns out, those folks are more interested in whatever larp they can do to scam folks & get money. I’ve also noticed a big overlap of the crypto larpers and AI hype shills.
I’m not complaining though. The fact that they aren’t here makes it a better experience for me.
Their model is literally as good as Opus and while it’s the same price you get a higher rate limit (supposedly). I wouldn’t call it overpriced. But yes Z-ai models are also very good and that’s the one I use personally for stuff. I’m okay if it screws up a bit on my personal projects.
Unfortunately I think CC subs can’t be used outside CC as of like a week ago but there’s so many good models out there. Kimi is cheaper and just as good.
TUIs were not meant to be written in fking JavaScript with react. “Vibe coding” can’t save you from shite engineering decisions. The only one I’ve used that doesn’t suck is crush (missing some features to be fair but doesn’t lag after 10m).
github.com/charmbracele...
A surprising amount of AI fear mongering can be boiled down to this simple trope.
See also:
PROGRAMMER: “I gave an LLM access to launch the nukes with no oversight”
LLM: “I launched the nukes”
PROGRAMMER: “Oh no how could you have done this?!? AI is so unsafe”
When is the talk? I'd love to swing by & listen :)
Hah I was reviewing a paper recently and had exactly this critique. They were like “our thing is better because of table 5” which was exclusively pass@16.
Then somewhere else they had a different benchmark with a single pass@1 score and it was legit identical to the baseline… smh
Stay tuned I’m gonna be getting up to some real shenanigans here shortly!
A computer terminal showing 8 AMD Instinct MI300X GPUs available in PyTorch
👀
Uhh guys… did they just build gastown into the model??????
www.kimi.com/blog/kimi-k2...
The prompt processing speed is also super slow on macOS once the context gets big but perhaps that’s because I was bleeding over into swap idk.
I used Q8 on my MacBook and despite being just on the edge of OOM it was able to write a very coherent C++ softmax implementation. Q6 also seemed to work just fine. But I couldn’t get it to really do any agent workflows because of the aformentioned bug.
I mean it won’t be any different. Lm studio has an mlx backend option and it also allows for the max size of the context window to be controlled. But there’s a bug which is impacting basically all runtimes w/ this model rn.
github.com/ml-explore/m...
You can either wait a few days for these patches to make their way to the main line release or build from source / use nightly builds.