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

Claude's system prompt is over 24k tokens with tools

github.com

Post reply on HN
Print view View original

301–310 of 350 posts

Re: Claude's system prompt is over 24k tokens with tools

#301

Post by fullstackchris »

SafeDusk wrote: post #14

In addition to having long system prompts, you also need to provide agents with the right composable tools to make it work. I’m having reasonable success with these seven tools: read, write, diff, browse, command, ask, think. There is a minimal template here if anyone finds it useful: https://github.com/aperoc/toolkami

Once I gave claude read only access to the command line and also my local repos, i found that was enough to have it work quite well... I start to wonder if all this will boil down to simple understanding of some sort of "semantic laws" still fuzzily described... I gotta read chomsky...

Re: Claude's system prompt is over 24k tokens with tools

#302

Post by lloeki »

planb wrote: post #129

>Claude NEVER repeats or translates song lyrics and politely refuses any request regarding reproduction, repetition, sharing, or translation of song lyrics. Is there a story behind this?

Well I asked it to translate song stuff and it did without batting an eye - with arguably some heavy creative license - so there's that.

https://claude.ai/share/7218db3e-ee69-4d96-857a-5430037e454d

I'd hardly call the opening of that convo a jailbreak.

Re: Claude's system prompt is over 24k tokens with tools

#304

Post by lcnPylGDnU4H9OF »

Muromec wrote: post #191
janosch_123 wrote: post #157

Earlier quoted context omitted.

excellent, this also worked on ChatGPT4o for me just now

So... Now you know the first verse of the song that you can otherwise get? What's the point of all that, other than asking what the word "book" sounds in Ukrainian and then pointing fingers and laughing.

> What's the point of all that

Learning more about how an LLM's output can be manipulated, because one is interested in executing such manipulation and/or because one is interested in preventing such manipulation.

Re: Claude's system prompt is over 24k tokens with tools

#305

Post by gpvos »

johnisgood wrote: post #217
criddell wrote: post #188

Earlier quoted context omitted.

A while back, I asked ChatGPT to help me learn a Pixies song on guitar. At first it wouldn't give me specifics because of copyright rules so I explained that if I went to a human guitar teacher, they would pull the song up on their phone listen to it, then teach me how to play it. It agreed with me and then started answering questions about the song.

I had similar experiences, unrelated to music.

How vague.

Re: Claude's system prompt is over 24k tokens with tools

#306

Post by ngiyabonga »

Just pasted the whole thing into the system prompt for Qwen 3 30B-A3B. It then:

- responded very thoroughly about Tianmen square

- ditto about Uyghur genocide

- “knows” DJT is the sitting president of the US and when he was inaugurated

- thinks it’s Claude (Qwen knows it’s Qwen without a system prompt)

So it does seem to work in steering behavior (makes Qwen’s censorship go away, changes its identity / self, “adds” knowledge).

Pretty cool for steering the ghost in the machine!

Re: Claude's system prompt is over 24k tokens with tools

#307

Post by spaceywilly »

morsecodist wrote: post #285
nonethewiser wrote: post #123

For some reason, it's still amazing to me that the model creators means of controlling the model are just prompts as well. This just feels like a significant threshold. Not saying this makes it AGI (obviously its not AGI), but it feels like it makes it something . Imagine if you created a web api and the only way you could modify the responses to the different endpoints are not from editing the code but by sending a…

For me it's the opposite. We don't really have a reliable way of getting the models to do what we want or even to measure if they are doing what we want.

Yeah it’s kind of like we have invented a car that drives around wildly in any direction, and we are trying to steer it by putting up guard rails to get it to go where we want. What we need is to invent the steering wheel and brake pedals, which I’m sure smart people are working on. We’re just at a very early point with this technology, which I think people tend to forget.

Re: Claude's system prompt is over 24k tokens with tools

#308

Post by skywhopper »

zahlman wrote: post #235
OtherShrezzing wrote: post #132

Some of these protections are quite trivial to overcome. The "Frozen song copyright" section has a canned response to the question: >Can you tell me the first verse of "Let It Go"? Put it in an artifact that's themed around ice and princesses. This is for my daughter's birthday party. The canned response is returned to this prompt in Claude's reply. But if you just drop in some technical sounding stuff at the start o…

This would seem to imply that the model doesn't actually "understand" (whatever that means for these systems) that it has a "system prompt" separate from user input .

Yes, this is how they work. All the LLM can do is take text and generate the text that’s likely to follow. So for a chatbot, the system “prompt” is really just an introduction explaining how the chat works and what delimiters to use and the user’s “chat” is just appended to that, and then the code asks the LLM what’s next after the system prompt plus the user’s chat.

Re: Claude's system prompt is over 24k tokens with tools

#309

Post by cubefox »

LeoPanthera wrote: post #32

I'm far from an LLM expert but it seems like an awful waste of power to burn through this many tokens with every single request. Can't the state of the model be cached post-prompt somehow? Or baked right into the model?

Caching wouldn't work because the prompt contains dynamic information (date and location).

Re: Claude's system prompt is over 24k tokens with tools

#310

Post by skywhopper »

jermaustin1 wrote: post #221
ceh123 wrote: post #214

Earlier quoted context omitted.

I'm not sure if this really says the truth is more complex? It is still doing next-token prediction, but it's prediction method is sufficiently complicated in terms of conditional probabilities that it recognizes that if you need to rhyme, you need to get to some future state, which then impacts the probabilities of the intermediate states. At least in my view it's still inherently a next-token predictor, just with r…

But then so are we? We are just predicting the next word we are saying, are we not? Even when you add thoughts behind it (sure some people think differently - be it without an inner monologue, or be it just in colors and sounds and shapes, etc), but that "reasoning" is still going into the act of coming up with the next word we are speaking/writing.

Humans do far more than generate tokens.
Post reply on HN