(openai.com)

1053 points atgctg | 3 comments | 11 Dec 25 18:04 UTC | HN request time: 0.001s | source

Show context

jumploops ◴[11 Dec 25 18:58 UTC] No.46235526[source]▶

> “a new knowledge cutoff of August 2025”

This (and the price increase) points to a new pretrained model under-the-hood.

GPT-5.1, in contrast, was allegedly using the same pretraining as GPT-4o.

1. redox99 ◴[12 Dec 25 02:04 UTC] No.46240048[source]▶

I think it's more likely to be the old base model checkpoint further trained on additional data.

replies(1): >>46242079 #

2. jumploops ◴[12 Dec 25 08:34 UTC] No.46242079[source]▶

Is that technically not a new pretrained model?

(Also not sure how that would work, but maybe I’ve missed a paper or two!)

replies(1): >>46242304 #

3. redox99 ◴[12 Dec 25 09:16 UTC] No.46242304[source]▶

I'd say for it to be called a new pretrained model, it'd need to be trained from scratch (like llama 1, 2, 3).

But it's just semantics.

GPT-5.2