_0ffh

Yeah, until you follow the link only to find out that the link does not say what Perplexity told you it does. This happened to me a frightful number of times.

Funny, my logging module for Python does the same (if requested with an optional keyword argument). It fetches the call stack, removes it's own functions, replaces Python library function chains with a single [...] and prepends it to the log output (together with usual date and time).

Quite often these advertisers are also doing this primarily to appease, or at least are triggered into action by pressure groups.

Nobody is getting "muted or silenced by mindless, heartless algorithms"!

They are getting muted or silenced by busybodies who try to control language, and are either in a position to install these algorithms, or are part of pressure mobs that aim to influence the former.

Hmm, I have read contradicting accounts.

They say the term did not originate from the steam condenser. Quite the opposite, it became deprecated because of the ambiguous meaning of steam condenser.

Yes. Also, while the model having a certain instruction once in it's context might count as the agent "knowing" about it, I'm almost certain that repeating those instructions - especially right at the position in the context where it counts - should almost certainly make a difference. Of course I can't know how much of a difference exactly, but that's what experiments are for.

I think this is a really excellent idea!

I'm almost sure that should at least lower the inclination of the model to try and "fix" the access problem, and I want to see this implemented and systematically evaluated.

I wish I could highlight this more than just with a vote and a reply, but I'll just have to be content with doing what I can here.

True, though the timeline was more the other way around. Error backprop was used as a method for training before using AD to automate the implementation work became the norm. Previously you had to write the backward pass by hand, now we use AD to derive it from the forward pass.

Easily circumvented: Generate a random seed in the shell and pass it to the core.

Well, there's essentially two major ways to use these models: Pair programming or fully autonomous fire-and-forget code generation. The second strategy needs essentially zero input, so the number of tokens you can blow is practically only limited by API speed.

Arrogant, pseudo-intellectual drivel.

That's what I though. I've got a harness that uses landlock. Might not be perfect, but should be good enough for almost all cases.

Inadvertent touching is fixable, ARM for example did it with the S bit (though on AArch64 it's slightly more complicated).

I regard it as a mistake of RISC-V. The flag register was invented for good reasons, and dropping it is a trade-off I personally do not think is worth the downside.

>absolute distain for flag registers

It's worst aspect maybe.

Damn, I remember I worked through the whole damn thing some years ago when I read this article for the first time, and it didn't hold up to it's claim. I was actually contemplating to do a write-up, but it wasn't really important to me so I didn't. Seeing the same article re-appear on HN for the second time now I really wish that I did. Unfortunately I've forgotten the details by now.

In "biometrics that match insightface bit for bit" - the "bit for bit" thing is also something I have encountered more than once.

Yes, stupid and lazy people will be stupid and lazy.

AI is just another tool they can use to be stupid and lazy while pretending to contribute.

Could have been put in a few short sentences, most of which would be wrong.

I know chashu is the pork, but I don't remember the broth type. Sure it wasn't miso, I think maybe salt or soy sauce type.

I've been to Japan - the first cheap buy-and-nuke-at-home ramen from the konbini around the corner I ate there completely blew me away (†)! It was decisively better than what I would realistically expect to get in a restaurant in the west. I don't know how they do it, but I learned never to underestimate the Japanese people’s dedication to their craft.

(†) For the curious, it was chāshū ramen.

Edit: To avoid confusion, I do not mean instant ramen, but a plastic bowl of chilled, pre-prepared take away ramen soup.

I don't see that, where do you read that?

Also, what kind of papers do they publish at all?

When I look at the recently (say, last 1-2 years) published practical improvements in the LM space, they're not even close to what the Chinese are publishing.

Yup, I remember early school. Two guys in class had a computer, one was a ZX81 the other (me) a C64. We're still friends! =)

This is... strange? I wonder what the reason might be?

Anyways, you've found someone. Or at least someone who used to.

Well, iff is a useful abbreviation. I use it in personal notes at least, and would use it more often if I thought it was more widely understood.

Oh, I assume they're innovating - it's what I meant with "doing new things".

But the word pioneer comes from French pionnier, literally “foot soldier”, a soldier who goes ahead to prepare the way.

If you don't publish you may be advancing, but you're not preparing anyone's way.

I don't need to. Those who get it get it. Apparently that doesn't include you, and I'm fine with that.

I'm not.

I'm afraid I'm even balking at the word "pioneering" in context with US frontier labs. They are probably doing a few new things, right, but they are not blazing any trails for others to follow along, the Chinese are.

Lookahead Sparse Attention should be playing a big role as well, as it dramatically slashes memory consumption.

They can only pay so much because of the expectations about the generated value. Those might turn out wrong, as any expectation, but value is very much in the picture.