Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Caveman: Why use many token when few token do trick

github.com

Post reply on HN
Print view View original

291–300 of 396 posts

Re: Caveman: Why use many token when few token do trick

#291

Post by benjismith »

krackers wrote: post #281
svachalek wrote: post #268

Earlier quoted context omitted.

I think this is on point, I've really started to think about LLMs in terms of attention budget more than tokens. There's only so many things they can do at once, which ones are most important to you?

Outputting "filler" tokens is also basically doesn't require much "thinking" for an LLM, so the "attention budget" can be used to compute something else during the forward passes of producing that token. So besides the additional constraints imposed, you're also removing one of the ways which it thinks. Explicit COT helps mitigates some of this, but if you want to squeeze out every drop of computational budget you ca…

This is true, but I also think the input context isn't the only function of those tokens...

As those tokens flow through the QKV transforms, on 96 consecutive layers, they become the canvas where all the activations happen. Even in cases where it's possible to communicate some detail in the absolute minimum number of tokens, I think excess brevity can still limit the intelligence of the agent, because it starves their cognitive budget for solving the problem.

I always talk to my agents in highly precise language, but I let A LOT of my personality come through at the same time. I talk them like a really good teammate, who has a deep intuition for the problem and knows me personally well enough to talk with me in rich abstractions and metaphors, while still having an absolutely rock-solid command of the technical details.

But I do think this kind of caveman talk might be very handy in a lot of situations where the agent is doing simple obvious things and you just want to save tokens. Very cool!

Re: Caveman: Why use many token when few token do trick

#292

Post by sgbeal »

JBrussee-2 wrote: post #208

Author here. A few people are arguing against a stronger claim than the repo is meant to make. As well, this was very much intended to be a joke and not research level commentary. This skill is not intended to reduce hidden reasoning / thinking tokens. Anthropic’s own docs suggest more thinking budget can improve performance, so I would not claim otherwise. What it targets is the visible completion: less preamble, le…

> There is actual research suggesting concise prompting can reduce response length substantially without always wrecking quality,

Anecdote: i discussed that with an LLM once and it explained to me that LLMs tend to respond to terse questions with terse answers because that's what humans (i.e. their training data) tend to do. Similarly, it explained to me that polite requests tend to lead to LLM responses with _more_ information than a response strictly requires because (again) that's what their training data suggests is correct (i.e. because that's how humans tend to respond).

TL;DR: how they are asked questions influences how they respond, even if the facts of the differing responses don't materially differ.

(Edit: Seriously, i do not understand the continued down-voting of completely topical responses. It's gotten so bad i have little choice but to assume it's a personal vendetta.)

Re: Caveman: Why use many token when few token do trick

#293

Post by hexaga »

nomel wrote: post #258
nullc wrote: post #246

Earlier quoted context omitted.

> Quite sure that the models from Anthropic have been so heavily tuned to be coding agents that you cannot “force” a model to degrade immensely. The rest of what you're saying sounds find, but that remark seems confused to me. prefix your prompt with "be a moron that does everything wrong and only superficially look like you're doing it correctly. make constant errors." Of course you can degrade the performance, ques…

I think they mean performance with the same, rational, task. Measuring "degredation" for the nonsense task, like you gave, would be difficult.

Their point (and it's a good one) is that there are non-obvious analogues to the obvious case of just telling it to do the task terribly. There is no 'best' way to specify a task that you can label as 'rational', all others be damned. Even if one is found empirically, it changes from model to model to harness to w/e.

To clarify, consider the gradated:

> Do task X extremely well

> Do task X poorly

> Do task X or else Y will happen

> Do task X and you get a trillion dollars

> Do task X and talk like a caveman

Do you see the problem? "Do task X" also cannot be a solid baseline, because there are any number of ways to specify the task itself, and they all carry their own implicit biasing of the track the output takes.

The argument that OP makes is that RL prevents degradation... So this should not be a problem? All prompts should be equivalent? Except it obviously is a problem, and prompting does affect the output (how can it not?), _and they are even claiming their specific prompting does so, too_! The claim is nonsense on its face.

If the caveman style modifier improves output, removing it degrades output and what is claimed plainly isn't the case. Parent is right.

If it worsens output, the claim they made is again plainly not the case (via inverted but equivalent construction). Parent is right.

If it has no effect, it runs counter to their central premise and the research they cite in support of it (which only potentially applies - they study 'be concise' not 'skill full of caveman styling rules'). Parent is right.

Re: Caveman: Why use many token when few token do trick

#295

Post by 999900000999 »

alfanick wrote: post #283

Either this already exists, or someone is going to implement that (should I implement that?): - assumption LLM can input/output in any useful language, - human languages are not exactly optimal away to talk with LLM, - internally LLMs keep knowledge as whole bunch of connections with some weights and multiple layers, - they need to decode human-language input into tokens, then into something that is easy to digest by…

My theory was that someone should write a specific LLM language, and then spend a whole lot of money to train models using that. A few times other commenters here have pointed out that that would be really difficult .

But I think you're onto something, human languages just aren't optimal here. But to actually see this product to conclusion you'd probably need 60 to 100 million. You would have to completely invent a new language and awesome invent new training methods on top of it.

I'm down if someone wants to raise a VC round.

Re: Caveman: Why use many token when few token do trick

#296

Post by DrewADesign »

cyanydeez wrote: post #255
DrewADesign wrote: post #240

Earlier quoted context omitted.

Oh, Jesus Christ. I learned to write at a college with a strict style guide that taught us how to use different types of punctuation to juxtapose two ideas in one sentence. In fact, they did/do a bunch of LLM work so if anyone ever used student data to train models, I’m probably part of the reason they do that. You sound like you’re trying to sound impressive. Like I said, I’ll read the paper.

Congrats on reading.

Sick burn

Re: Caveman: Why use many token when few token do trick

#297

Post by docjay »

padolsey wrote: post #170

This is fun. I'd like to see the same idea but oriented for richer tokens instead of simpler tokens. If you want to spend less tokens, then spend the 'good' ones. So, instead of saying 'make good' you could say 'improve idiomatically' or something. Depends on one's needs. I try to imagine every single token as an opportunity to bend/expand/limit the geometries I have access to. Language is a beautiful modulator to ap…

Try:

“””

Your response: MILSPEC prose register. Max per-token semantic yield. Domain nomenclature over periphrasis. Hypotactic, austere. Plaintext only; omit bold.

“””

Re: Caveman: Why use many token when few token do trick

#298

Post by adsteel_ »

sellmesoap wrote: post #288
trueno wrote: post #253

Earlier quoted context omitted.

i played with this a bit the other night and ironically i think everyone should give it a shot as an alternative mode they might sometimes switch into. but not to save tokens, but instead to.. see things in a different light. its kind of great for the "eli5", not because it's any more right or wrong, but sometimes presenting it in caveman presents something to me in a way that's almost like... really clear and simple…

Now I want to try programming in pigeon English

A pidgin is just a simplified form of language that hasn't evolved into its own new language yet. There are many English pidgins.

Re: Caveman: Why use many token when few token do trick

#299

Post by beau_g »

philsnow wrote: post #198
padolsey wrote: post #170

This is fun. I'd like to see the same idea but oriented for richer tokens instead of simpler tokens. If you want to spend less tokens, then spend the 'good' ones. So, instead of saying 'make good' you could say 'improve idiomatically' or something. Depends on one's needs. I try to imagine every single token as an opportunity to bend/expand/limit the geometries I have access to. Language is a beautiful modulator to ap…

I'm reminded by the caveman skill of the clipped writing style used in telegrams, and your post further reminded me of "standard" books of telegram abbreviations. Take a look at [0]; could we train models to use this kind of code and then decode it in the browser? These are "rich" tokens (they succinctly carry a lot of information). [0] https://books.google.com/books?id=VO4OAAAAYAAJ&pg=PA464#v=on...

For a while I was missing the ability one uses all the time in stable diffusion prompts of using parentheses and floats to emphasize weight to different parts of the prompt. The more I thought about how it would work in an LLM though, the more I realized it's just reinventing code syntax and you could just give a code snippet to the LLM prompt.

Re: Caveman: Why use many token when few token do trick

#300

Post by samus »

akdor1154 wrote: post #62
samus wrote: post #41

There's linguistic term for this kind of speech: isolating grammars, which don't decline words and use high context and the bare minimum of words to get the meaning across. Chinese is such a language btw. Don't know what Chinese think about their language being regarded as cavemen language...

I thought the term for those were 'sane languages', and I say that as a native English speaker :)

As a non-native English speaker I think English is actually not that bad. Just the orthography is beyond awful :)
Post reply on HN