0-_-0

6 ms·

1.
by 0-_-0 10d ago
Instruct the AI to output an .MD file that can be used to reconstruct the code
2.
by 0-_-0 22d ago
Why didn't Russia just use local models?
3.
by 0-_-0 1mo ago
You're hallucinating
4.
by 0-_-0 1mo ago
Affected by what exactly?
5.
by 0-_-0 1mo ago
The kv cache is not discretised, only the output tokens.
6.
by 0-_-0 1mo ago
Reasoning is already taking place in latent space, the CoT is just a sideeffect
7.
by 0-_-0 1mo ago
I wonder if you could prune the kv cache as well
8.
by 0-_-0 1mo ago
Same is true for 1
9.
by 0-_-0 1mo ago
Right, so it was sealed only in the biological sense, which makes sense since there needs to be an inflow of energy
10.
by 0-_-0 1mo ago
No but they could over 5 million years
11.
by 0-_-0 1mo ago
The information that was missing from my life: For the Average Man, gaining an inch of height is worth losing roughly $25k in annual salary. Meanwhile, an inch of pecker is roughly offset by reducing annual salary by $128k.
12.
by 0-_-0 2mo ago
You would need to hide the existence of the original profile while in the decoy profile for this to work, which GrapheneOS considers too complex to implement
13.
by 0-_-0 2mo ago
It's not a benchmark, it's a meme
14.
by 0-_-0 2mo ago
They want to be power efficient, while Nvidia doesn't. That means a much higher cost (larger chip area) for the same performance.
15.
by 0-_-0 2mo ago
Yes, but all subsequent tokens do depend on them, so they are still being charged for
16.
by 0-_-0 2mo ago
Don't you jump the highest with com_maxfps 333 anyway? Or did Quake Live change this from Quake 3?
17.
by 0-_-0 2mo ago
You can only see satellites during twilight when they can reflect sunlight. Don't panic.
18.
by 0-_-0 2mo ago
In that article Grok shows highest neutrality
19.
by 0-_-0 2mo ago
This made me wonder about a 3D printer alternative that builds things by folding a thin sheet of metal into arbitrary shapes instead of extruding filament.
20.
by 0-_-0 2mo ago
Even if you had all that you would get completely different weights out at the end, and you also don't have the resources to "compile" an LLM because the compilation can cost $100M. If you were given the training data but not
21.
by 0-_-0 2mo ago
Source code makes it easier to modify a binary, training data doesn't make it easier to modify a trained network.
22.
by 0-_-0 2mo ago
If we can launch 1M satellites, how many telescopes can we launch?
23.
by 0-_-0 2mo ago
And what are they reflecting when they are in Earth's shadow?
24.
by 0-_-0 3mo ago
32 bits vs 4 bits it looks like
25.
by 0-_-0 3mo ago
Browser fingerprinting can already pinpoint you exactly. We should focus on that.
26.
by 0-_-0 3mo ago
That math makes no sense
27.
by 0-_-0 3mo ago
These never ever accomplish anything
28.
by 0-_-0 3mo ago
Not if it's mass produced
29.
by 0-_-0 3mo ago
No, it means the motor is smaller and it can be put into the wheel
30.
by 0-_-0 3mo ago
And this is just fraud that was done with incompetence, so easily caught. How much is done competently?

More ›