We learned something about how LLMs work with this… its like a bunch of paintings were chopped up into pixels to use to make other paintings. No one knew it was possible to break the model and have it spit out the pixels of a single painting in order.
I wonder if diffusion models have some other wierd querks we have yet to discover
I’m not an expert, but I would say that it is going to be less likely for a diffusion model to spit out training data in a completely intact way. The way that LLMs versus diffusion models work are very different.
LLMs work by predicting the next statistically likely token, they take all of the previous text, then predict what the next token will be based on that. So, if you can trick it into a state where the next subsequent tokens are something verbatim from training data, then that’s what you get.
Diffusion models work by taking a randomly generated latent, combining it with the CLIP interpretation of the user’s prompt, then trying to turn the randomly generated information into a new latent which the VAE will then decode into something a human can see, because the latents the model is dealing with are meaningless numbers to humans.
In other words, there’s a lot more randomness to deal with in a diffusion model. You could probably get a specific source image back if you specially crafted a latent and a prompt, which one guy did do by basically running img2img on a specific image that was in the training set and giving it a prompt to spit the same image out again. But that required having the original image in the first place, so it’s not really a weakness in the same way this was for GPT.
But the fact is the LLM was able to spit out the training data. This means that anything in the training data isn’t just copied into the training dataset, allegedly under fair use as research, but also copied into the LLM as part of an active commercial product. Sure, the LLM might break it down and store the components separately, but if an LLM can reassemble it and spit out the original copyrighted work then how is that different from how a photocopier breaks down the image scanned from a piece of paper then reassembles it into instructions for its printer?
The technology of compression a diffusion model would have to achieve to realistically (not too lossily) store “the training data” would be more valuable than the entirety of the machine learning field right now.
IIRC based on the source paper the “verbatim” text is common stuff like legal boilerplate, shared code snippets, book jacket blurbs, alphabetical lists of countries, and other text repeated countless times across the web. It’s the text equivalent of DALL-E “memorizing” a meme template or a stock image – it doesn’t mean all or even most of the training data is stored within the model, just that certain pieces of highly duplicated data have ascended to the level of concept and can be reproduced under unusual circumstances.
Welcome to the wild West of American data privacy laws. Companies do whatever the fuck they want with whatever data they can beg borrow or steal and then lie about it when regulators come calling.
If I take a certain proportion of Thyme and mix it with another certain proportion of Basil the recipe for the resultant spice mix can be a sensitive, proprietary business secret despite there being nothing inherently sensitive about the ingredients themselves.
How can the training data be sensitive, if noone ever agreed to give their sensitive data to OpenAI?
Exactly this. And how can an AI which “doesn’t have the source material” in its database be able to recall such information?
Model is the right term instead of database.
We learned something about how LLMs work with this… its like a bunch of paintings were chopped up into pixels to use to make other paintings. No one knew it was possible to break the model and have it spit out the pixels of a single painting in order.
I wonder if diffusion models have some other wierd querks we have yet to discover
I’m not an expert, but I would say that it is going to be less likely for a diffusion model to spit out training data in a completely intact way. The way that LLMs versus diffusion models work are very different.
LLMs work by predicting the next statistically likely token, they take all of the previous text, then predict what the next token will be based on that. So, if you can trick it into a state where the next subsequent tokens are something verbatim from training data, then that’s what you get.
Diffusion models work by taking a randomly generated latent, combining it with the CLIP interpretation of the user’s prompt, then trying to turn the randomly generated information into a new latent which the VAE will then decode into something a human can see, because the latents the model is dealing with are meaningless numbers to humans.
In other words, there’s a lot more randomness to deal with in a diffusion model. You could probably get a specific source image back if you specially crafted a latent and a prompt, which one guy did do by basically running img2img on a specific image that was in the training set and giving it a prompt to spit the same image out again. But that required having the original image in the first place, so it’s not really a weakness in the same way this was for GPT.
But the fact is the LLM was able to spit out the training data. This means that anything in the training data isn’t just copied into the training dataset, allegedly under fair use as research, but also copied into the LLM as part of an active commercial product. Sure, the LLM might break it down and store the components separately, but if an LLM can reassemble it and spit out the original copyrighted work then how is that different from how a photocopier breaks down the image scanned from a piece of paper then reassembles it into instructions for its printer?
It’s not copied as is, thing is a bit more complicated as was already pointed out
The technology of compression a diffusion model would have to achieve to realistically (not too lossily) store “the training data” would be more valuable than the entirety of the machine learning field right now.
Overfitting.
IIRC based on the source paper the “verbatim” text is common stuff like legal boilerplate, shared code snippets, book jacket blurbs, alphabetical lists of countries, and other text repeated countless times across the web. It’s the text equivalent of DALL-E “memorizing” a meme template or a stock image – it doesn’t mean all or even most of the training data is stored within the model, just that certain pieces of highly duplicated data have ascended to the level of concept and can be reproduced under unusual circumstances.
Welcome to the wild West of American data privacy laws. Companies do whatever the fuck they want with whatever data they can beg borrow or steal and then lie about it when regulators come calling.
If I take a certain proportion of Thyme and mix it with another certain proportion of Basil the recipe for the resultant spice mix can be a sensitive, proprietary business secret despite there being nothing inherently sensitive about the ingredients themselves.
if i stole my neighbours thyme and basil out of their garden, mix them into certain proportions, the resulting spice mix would still be stolen.
all the leaked training data we have seen has been publicly available information so I don’t see the relevance unless you’re just trying to be mad