News for hackers. Stuff that matters.
Re:Tencent Releases and Open-Sources Tencent Hy4 Preview by demibabs on Saturday August 29, 2026 @08:35PM (#49494575)
No one’s talking about how good the final product is.

Edit: someone else commented that as I was typing this, lol.

Re:Tencent Releases and Open-Sources Tencent Hy4 Preview by blackhaz on Sunday August 30, 2026 @03:29AM (#49496501)
I wonder, do we need a new benchmark? There's quite a bit of feedback data floating around about pelicans on bicycles already.
Re:Tencent Releases and Open-Sources Tencent Hy4 Preview by ziofill on Sunday August 30, 2026 @05:05AM (#49496980)
That’s a fair question, but it seems that it’s not yet necessary. See here

https://dylancastillo.co/posts/pelicanmaxxing.html

https://simonwillison.net/2026/Jul/22/

Re:Tencent Releases and Open-Sources Tencent Hy4 Preview by stymaar on Sunday August 30, 2026 @11:50AM (#49499774)
I don't think this argument is a good one though, as it would be quite natural for a lab rhat want to macimize the performance of their model on the pelican bench to train it for “text-to-svg simple image generation” rather than just “pelicans on bicycle”.
Re:Tencent Releases and Open-Sources Tencent Hy4 Preview by Anon1096 on Sunday August 30, 2026 @01:51PM (#49501018)
That is the point though, if labs are maximizing svg image generation capabilities it is a very good thing. That's a general skill that is useful. So assuming they aren't specifically maximizing pelican bicycle svgs (and it doesn't look like they are) then incentives are aligned that the "benchmark" is measuring a general desirable capability.
Re:Tencent Releases and Open-Sources Tencent Hy4 Preview by flexagoon on Sunday August 30, 2026 @04:57PM (#49502725)
Re:Tencent Releases and Open-Sources Tencent Hy4 Preview by simonw on Sunday August 30, 2026 @08:19PM (#49504238)
Gemini have done exactly that.

(I doubt it's because of my stupid benchmark, though!)