4 ms·
My theory is that they're using a fine-tuned version of GPT-4-turbo in ChatGPT to fix the issue with everyone saying it was lazy. They fine-tuned it using respo
by atleastoptimal 3y ago
My theory is that they're using a fine-tuned version of GPT-4-turbo in ChatGPT to fix the issue with everyone saying it was lazy. They fine-tuned it using responses from GPT-4.5-turbo, which they have internally. This synthetic data was probably based on a suite of questions which included self-identification, leading to it releasing its model name in the training data, and the new ChatGPT less lazy model is adopting the name via osmotic data leakage.
- arthurcolle 3y agoGreat theory, this makes a lot of sense. It looks like they already RLHF'ed (I keep using this as a verb, but maybe its really just instruction tuning?) this behavior out of ChatGPT! https://news.ycombinator.com/item?id=38677025 https://news.ycombinator.com/item?id=38677025 Really fast, I wonder how this works. Anyone have any ideas / knowledge on how they deploy little incremental fixes to exploited jailbreaks, etc?)
- azeirah 3y agoNot sure about this scenario precisely but I believe when it comes to filtering certain responses they have a separate model running in a similar manner to what meta recently released. I believe it's called llama purple? https://ai.meta.com/llama/purple-llama/ https://ai.meta.com/llama/purple-llama/
- krisoft 3y ago> Anyone have any ideas / knowledge on how they deploy little incremental fixes to exploited jailbreaks, etc? LoRa[1] would be my guess. For detailed explanation I recommend the paper. But the short explanation is that it is a trick which lets you train a smaller, lower dimensional model which when you add to the original model it gets you the result you want. 1: https://arxiv.org/abs/2106.09685 https://arxiv.org/abs/2106.09685
- arthurcolle 3y agois this another expert? can they hot swap in experts and remove others?
- jug 3y agoYes I still think it’s involved because hallucinations don’t use to be this coherent across the user base.
- toastmaster11 3y agoBut why would it be gpt-4.5-turbo, as opposed to just gpt-4.5? Are they just going to abandon past naming conventions? If they already had 4.5 and then did something else to make it “turbo” why not release 4.5 on the coattails of Gemini?
- zamadatix 3y agoI don't think there would be pressure to rush a new model out immediately when Gemini Ultra (the one which is supposed to be better than the current product) doesn't even have a release date yet.
- toastmaster11 3y agoI’m asking, why is it 4.5-turbo when 4.5 isn’t released yet?
- idopmstuff 3y agoOr as they updated its knowledge cutoff, it picked up references to gpt-4.5-turbo. They wouldn't have to be internal - plenty of people have speculated about the possible release of a model with that specific name.
- atleastoptimal 3y agoWhy would it be deceived that it is a model that hasn't been released yet though? There would be more explicit references to gpt-4-turbo in the general internet training data