b--l

5 ms·

1.
by b--l 12d ago
Agreed--if you can catch them of course.
2.
by b--l 25d ago
I need to stop using this website.
3.
by b--l 2mo ago
It's one of claude models. We've seen it a lot so it' become extremely grating to us. And it's always the same speech patterns. I really can't stand it. I don't even think it's intentional from anthropic.
4.
by b--l 3mo ago
Codex is one of the most infamous examples of slopware. Just having the window unhidden on my mac will cause it to use 100% of the GPU displaying the spinner message. THE SPINNER MESSAGE CAUSES 100% GPU USAGE ON AN MBP M5!! So any time you&
5.
by b--l 3mo ago
It's because they use gpt-5.5-xhigh (the money making* model) to build it. (*for them)
6.
by b--l 3mo ago
I doubt it. It's a shit a model.
7.
by b--l 3mo ago
The constitutional right to bare LLMs.
8.
by b--l 3mo ago
Ooohh yeah. I forgot completely about that naked shameless bribe. That one was even more overt than the plane.
9.
by b--l 3mo ago
The last Opus models were money models--as in designed to make them money.
10.
by b--l 3mo ago
It will be interesting to see. If you look at the GDPR fines record (well at least the last time I did) spain stands out for its frequency and size of fine.
11.
by b--l 3mo ago
`qwen-35b-3a` is really garbage (I know well because it's what I run locally). What quant are you running? Would people really pay $6 a month for it even if unlimited? That said, nice looking site and wish you the best for getting it o
12.
by b--l 3mo ago
A huge win for china. No one is going to forget this. Every single developer will remember this weekend and be pissed off. I'm sure it's a huge wakeup call for any dev that was asleep and not bothering with open source models.
13.
by b--l 3mo ago
There is not a single chance this would have happened under that admin. Not one single chance.
14.
by b--l 3mo ago
I found it tripped in most laughable situations by mere were words that could be related in some way to hacking but are in common use in programming. I would have to go back, examine my prompt for word that could be use in another context a
15.
Goody-2: Safer than fable, similar capability(goody2.ai)
4 points by b--l 3mo ago|0 comments
16.
by b--l 4mo ago
Just speculating but I "feel" 4.7 was post-trained using more synthetic techniques. The way it writes for one thing, it's "personality", is less human and more fatiguing-AI-slop like.
17.
by b--l 4mo ago
Thank you for the gold kind stranger.
18.
by b--l 5mo ago
If you assume he is being blackmailed then it makes sense. On any other level it does not.
19.
by b--l 5mo ago
grok-4.1-fast is the the number 2 model on this benchmark. ~~If you've used this model in real life to do any sort of programming, and have seen its output, you would know that there is something VERY wrong with your benchmark.~~ Edit:
20.
by b--l 5mo ago
If we learned anything from the code leak is that they essentially do not know what is in the blackbox of the code for that 500k line mass. So that's plausible.
21.
by b--l 5mo ago
"but most of our samples are from the last 2 months." There's your major issue. That's well within the brutal quantization window.