[liorn]

I asked it about Tailwind CSS (since I had problems with Claude not aware of Tailwind 4):

> Which version of tailwind css do you know?

> I have knowledge of Tailwind CSS up to version 3.4, which was the latest stable version as of my knowledge cutoff in January 2025.

[threeducks]

> Which version of tailwind css do you know?

LLMs can not reliably tell whether they know or don't know something. If they did, we would not have to deal with hallucinations.

[redman25]

They can if they've been post trained on what they know and don't know. The LLM can first been given questions to test its knowledge and if the model returns a wrong answer, it can be given a new training example with an "I don't know" response.

[dingnuts]

Oh that's a great idea, just do that for every question the LLM doesn't know the answer to!

That's.. how many questions? Maybe if one model generates all possible questions then

[nicce]

We should use the correct term: to not have to deal with bullshit.

[dudeinjapan]

I think “confabulation” is the best term.

“Hallucination” is seeing/saying something that a sober person clearly knows is not supposed to be there, e.g. “The Vice President under Nixon was Oscar the Grouch.”

Harry Frankfurt defines “bullshitting” as lying to persuade without regard to the truth. (A certain current US president does this profusely and masterfully.)

“Confabulation” is filling the unknown parts of a statement or story with bits that sound as-if they could be true, i.e. they make sense within the context, but are not actually true. People with dementia (e.g. a certain previous US president) will do this unintentionally. Whereas the bullshitter generally knows their bullshit to be false and is intentionally deceiving out of self-interest, confabulation (like hallucination) can simply be the consequence of impaired mental capacity.

[nicce]

I think the Frankfurt definition is a bit off.

E.g. from the paper ChatGPT is bullshit [1],

> Frankfurt understands bullshit to be characterized not by an intent to deceive but instead by a reckless disregard for the truth.

That is different than defining "bullshitting" as lying. I agree that "confabulation" could otherwise be more accurate. But with previous definition they are kinda synonyms? And "reckless disregard for the truth" may hit closer. The paper has more direct quotes about the term.

[1] https://link.springer.com/article/10.1007/s10676-024-09775-5

[dudeinjapan]

You're right. It's "intent to persuade with a reckless disregard for the truth." But even by this definition, LLMs are not (as far as we know) trying to persuade us of anything, beyond the extent that persuasion is a natural/structural feature of all language.

[SparkyMcUnicorn]

Interesting. It's claiming different knowledge cutoff dates depending on the question asked.

"Who is president?" gives a "April 2024" date.

[ethbr1]

Question for HN: how are content timestamps encoded during training?

[cma]

Claude 4's system prompt was published and contains:

"Claude’s reliable knowledge cutoff date - the date past which it cannot answer questions reliably - is the end of January 2025. It answers all questions the way a highly informed individual in January 2025 would if they were talking to someone from {{currentDateTime}}, "

https://docs.anthropic.com/en/release-notes/system-prompts#m...

[polynomial]

I thought best guesses were that Claude's system prompt ran to tens of thousands of tokens, with figures like 30,000 tokens being bandied about.

But the documentation page linked here doesn't bear that out. In fact the Claude 3.7 system prompt on this page clocks in at significantly less than 4,000 tokens.

[tough]

they arent.

a model learns words or tokens more pedantically but has no sense of time nor cant track dates

[svachalek]

Yup. Either the system prompt includes a date it can parrot, or it doesn't and the LLM will just hallucinate one as needed. Looks like it's the latter case here.

[manmal]

Technically they don’t, but OpenAI must be injecting the current date and time into the system prompt, and Gemini just does a web search for the time when asked.

[tough]

right but that's system prompting / in context

not really -trained- into the weights.

the point is you can't ask a model what's his training cut off date and expect a reliable answer from the weights itself.

closer you could do is have a bench with -timed- questions that could only know if had been trained for that, and you'd had to deal with hallucinations vs correctness etc

just not what llm's are made for, RAG solves this tho

[stingraycharles]

What would the benefits be of actual time concepts being trained into the weights? Isn’t just tokenizing the dates and including those as normal enough to yield benefits?

E.g. it probably has a pretty good understanding between “second world war” and the time period it lasted. Or are you talking about the relation between “current wall clock time” and questions being asked?

[tough]

OpenAI injects a lot of stuff, your name, sub status, recent threads, memory, etc

sometimes its interesting to peek up under the network tab on dev tools

[Tokumei-no-hito]

strange they would do that client side

[dawnerd]

I did the same recently with copilot and it of course lied and said it knew about v4. Hard to trust any of them.

[PeterStuer]

Did you try giving it the relevant parts of the tailwind 4 documentation in the prompt context?