Re:Qwen4-Max
by andy_ppp on Monday November 16, 2026 @04:19AM
(#50515618)
He can do whatever he wants, I personally think they are quite informative
[ Parent
| on HN ]
Is it informative? I really don't get what signal you're picking out of a bunch of similar looking drawings.
Re:Qwen4-Max
by anentropic on Monday November 16, 2026 @05:41AM
(#50516126)
It's interesting to see which model + thinking strength gets the details right, such as legs on opposite sides of the frame, a coherent bicycle etc
Some people are determined not to understand I think, but if they don't find it useful no worries, there's just no need to have a temper tantrum about what other people post on here especially when the majority (it's near the top for a reason) find it useful. Just don't worry about it - you can even hide whole threads on there if you don't like them, I use the feature all the time rather than taking the time to comment negatively!
Re:Qwen4-Max
by chrisweekly on Monday November 16, 2026 @08:01AM
(#50517151)
what "temper tantrum"?
I find it amusing that repetitive pelican posting is fine and folks should just minimize and move on, but responding with an opinion that pelicans are boring/not valuable/engagement farming is "temper tantrum". Why not apply the same standards to both?
Isn't it a bit arbitrary though?
The original concept of this "benchmark" was to see how the model performs on a reasonably complex task that probably isn't in the training data.
Even if they don't fine-tune for this specifically, there are now surely tons of examples of people asking for a pelican on a bike in the training data, and making an SVG isn't that complex of a task for current gen models.
Even with the same model at the same thinking level, you might get different quality out of the same prompt on different runs. If you want to use this as an eval, you would really want to generate 100 of them, and see how often they're anatomically/mechanically accurate.