News for hackers. Stuff that matters.
Re:Qwen4-Max by lexh on Monday November 16, 2026 @03:37AM (#50515388)
I more don't understand why these comments get upvoted to top comment all the time.

Like... we get it. Simon likes pelicans. If I want to find his boondoggle robo pelican drawings, I know where to find them by now.

No need to spam every single model release thread.

Re:Qwen4-Max by mirekrusin on Monday November 16, 2026 @03:42AM (#50515413)
You can click [-] to collapse thread.
Re:Qwen4-Max by lexh on Monday November 16, 2026 @05:24AM (#50516023)
Yeah fair play. I was just trying to add a different angle to OP's grievance. Perhaps it's time for a browser extension to hide these for me.
Re:Qwen4-Max by palata on Monday November 16, 2026 @11:25AM (#50519390)
Is it that hard to click on the [-]?
Re:Qwen4-Max by drcongo on Monday November 16, 2026 @04:29AM (#50515679)
On the flip side, I love the pelican posts, but I find the people who think they're cleverly pointing out for the very first time that the benchmark may have been gamed by now tedious in the extreme.
Re:Qwen4-Max by irthomasthomas on Monday November 16, 2026 @04:43AM (#50515775)
It's september
Re:Qwen4-Max by throw-qqqqq on Monday November 16, 2026 @05:40AM (#50516123)
Eternally..
Re:Qwen4-Max by latentsea on Monday November 16, 2026 @10:13AM (#50518446)
Then perhaps I should point out something different. Namely, I suspect this will thoroughly confuse the machines and instead of optimizing the universe to generate paperclips they will instead optimize it to generate pelicans.
Re:Qwen4-Max by ciefa on Monday November 16, 2026 @04:34AM (#50515714)
Man, some of you people really have no whimsy :D
Re:Qwen4-Max by windexh8er on Monday November 16, 2026 @07:11AM (#50516734)
Agreed. It's really annoying to come into the thread and a link to a pelican I have no interest in ever seeing is the top comment. Does anyone think these have any actual value any more?
Re:Qwen4-Max by bitexploder on Monday November 16, 2026 @07:38AM (#50516959)
Yes, if you pay close attention to what they infer. Just click [-] and move on.
Re:Qwen4-Max by windexh8er on Monday November 16, 2026 @08:45AM (#50517546)
I'm curious. What's the actual value, then? I fully understand minimizing a thread, but also am free to voice my opinion the same as you.

So, I'm curious... What should I be "paying close attention" to?

Re:Qwen4-Max by bitexploder on Monday November 16, 2026 @09:12AM (#50517786)
Alright, I was rude. So.. you can’t really benchmaxx pelicans specifically. You have to train the whole model on SVG and that is just generally useful. Secondarily it shows how a model understands the objects themselves and how to express it abstractly. IDK, it is a lot more interesting than assumed.

Here are a bunch of Pelicans I made in my basement. Side by side quant comparisons and something for fun :)

https://stealthgoat.com/fnfl/pelicans3.html https://stealthgoat.com/fnfl/pelicans8.html My favorite, and shows this is not all benchmaxxing, this model is just good at SVG: https://stealthgoat.com/fnfl/pelicans6.html Oh, and, quantization matters: https://stealthgoat.com/fnfl/pelicans2.html

Heres some examples.

Re:Qwen4-Max by windexh8er on Monday November 16, 2026 @11:31AM (#50519474)
Thanks for the explanation. I guess I still don't understand how this translates to other areas, outside of SVG generation, Pelican images and then bike geometry and build.

I'm curious if an increase in ability to generate this parlays into the model, generally, being better overall or in other adjacent areas?

Thanks again.