How outdated information hides in LLM token generation probabilities

Re:How outdated information hides in LLM token generation pr... by freehorse on Sunday January 12, 2025 @04:26AM (#42672320)

Does it search the internet for that? I assume so because else claiming how often something is cited does not make sense, but would be interesting to know surely. Even gpt4o mini with kagi gets it right with search enabled (and wrong without search enabled - tried over a few times to make sure).

[ Parent | on HN ]

Re:How outdated information hides in LLM token generation pr... by sd9 on Sunday January 12, 2025 @05:05AM (#42672472)

I don’t think the public o1 can search the internet yet, unlike 4o. In principle it could know that something is more commonly cited based on its training data. But it could also just be hallucinating.

[ Parent | on HN ]

Re:How outdated information hides in LLM token generation pr... by freehorse on Sunday January 12, 2025 @07:54AM (#42673228)

> it could know that something is more commonly cited based on its training data

No there is no such concept or way to do something like that. LLMs do not have such kind of meta-knowledge over their training data or weights. But there could be explicit mentions about this on their training data and they could pick on that and that is probably the simplest explanation.

[ Parent | on HN ]

Re:How outdated information hides in LLM token generation pr... by og_kalu on Sunday January 12, 2025 @11:12AM (#42674556)

>LLMs do not have such kind of meta-knowledge over their training data or weights.

Not sure this is a claim that can be confidently made.

https://arxiv.org/abs/2309.00667

https://x.com/flowersslop/status/1873115669568311727?t=eBMbK...

[ Parent | on HN ]

Re:How outdated information hides in LLM token generation pr... by diggan on Sunday January 12, 2025 @06:40AM (#42672861)

> In principle it could know that something is more commonly cited based on its training data

Could it? Without explicit training for that, how would it be expected to know it has to be able to count occurrences of something?

[ Parent | on HN ]

Re:How outdated information hides in LLM token generation pr... by sd9 on Sunday January 12, 2025 @07:29AM (#42673099)

I think it would be more vibes based - commonly occurring things would be reinforced more in the weights. Rather than it explicitly counting the number of occurrences.

[ Parent | on HN ]

Re:How outdated information hides in LLM token generation pr... by diggan on Sunday January 12, 2025 @09:03AM (#42673594)

So the probabilities would be skewed towards something, but unless the model could somehow count/infer its own weights, I don't see how it could "introspect" to see if something is more common than something else.

[ Parent | on HN ]

Re:How outdated information hides in LLM token generation pr... by asl2D on Sunday January 12, 2025 @04:31AM (#42672343)

Can the claim about citation frequency be just an answer pattern and not model's exact reasoning?

[ Parent | on HN ]

Re:How outdated information hides in LLM token generation pr... by freehorse on Sunday January 12, 2025 @04:42AM (#42672378)

Yeah there could be parts of the training set with 1611 being explicitly called the official and 1622 being explicitly called the most common answer. But it can also have access to search results directly I think. Is there a way to know if it does or not?

[ Parent | on HN ]