accountnum

7 ms·

1.
by accountnum 2y ago
It's not a problem, because the point at which we are in the logarithmic curve is the only thing that matters. No one in their right mind ever expected anything linear, because that would imply that creating a perfect oracle is possibl
2.
by accountnum 2y ago
I'm going to simply address what I think are your main points here. There is nowhere that an LLM stores all possible outputs. Causality can trivially be represented by sampling by including the ordering of events, which you also implic
3.
by accountnum 2y ago
Again, you're stretching definitions into meaninglessness. The way you are using "sampling" and "distribution" here applies to any system processing any information. Yes, humans as well. I can trivially define the e
4.
by accountnum 2y ago
They're not sampling from prior conversations. The model constructs abstracted representations of the domain-specific reasoning traces. Then it applies these reasoning traces in various combinations to solve unseen problems. If you wan
5.
by accountnum 2y ago
My point isn't that the model falls for gender stereotypes, but that it falls for thinking that it needs to solve the unmodified riddle. Humans fail at the original because they expect doctors to be male and miss crucial information be
6.
by accountnum 2y ago
Recognizing that it is a riddle isn't impressive, true. But the duration of its reasoning is irrelevant, since the riddle works on misdirection. As I keep saying here, give someone uninitiated the 7 wives with 7 bags going (or not) to
7.
by accountnum 2y ago
The trick with the 7 wives and 7 bags and so on is that no long reasoning is required. You just have to notice one part of the question that invalidates the rest and not shortcut to doing arithmetic because it looks like an arithmetic probl
8.
by accountnum 2y ago
No, that's not a conclusion we can draw, because there is nothing much more to do than memorize the answer to this specific trick question. That's why it's a trick question, it goes against expectations and therefore the gene
9.
by accountnum 2y ago
It hasn't read that riddle because it is a modified version. The model would in fact solve this trivially if it _didn't_ see the original in its training. That's the entire trick.
10.
by accountnum 2y ago
It literally is a riddle, just as the original one was, because it tries to use your expectations of the world against you. The entire point of the original, which a lot of people fell for, was to expose expectations of gender roles leading
11.
by accountnum 2y ago
No, it's necessary to either know that it's a trick question or to have a feeling that it is based on context. The entire point of a question like that is to trick your understanding. You're tricking the model because it has
12.
by accountnum 2y ago
No, you're not. Are you genuinely trying to suggest that LLMs, which can: - Construct arbitrary text that isn't just grammatically but semantically coherent - Derive intent, subtle intent, from user queries and responses - Emulate
13.
by accountnum 2y ago
A weakness of the current models in some domains considered useful, yes - but not a fundamental limitation of the architecture. I see no consensus on the latter whatsoever. The ARC challenge tests spatial reasoning, something we humans are
14.
by accountnum 2y ago
> One reason why just blathering on endlessly... First of all, I would urge you to stop arbitrarily using negative words to make an argument. Saying that LLMs are "blathering" is equivalent to saying you and I are "smackin
15.
by accountnum 2y ago
You seem to repeatedly insist that hidden computation is a distinction of any relevance whatsoever. First of all, your understanding of the architecture itself is mistaken. A transformer can iterate endlessly because each token it produces