3 ms·
>You are never going to get the exact book word-for-word using LLM. This is pretty much the exact claim of a NYT lawsuit against OpenAI. "One example: Bing Ch
by lbrito 5mo ago
>You are never going to get the exact book word-for-word using LLM.
This is pretty much the exact claim of a NYT lawsuit against OpenAI.
"One example: Bing Chat copied all but two of the first 396 words of its 2023 article “The Secrets Hamas knew about Israel’s Military.” An exhibit showed 100 other situations in which OpenAI’s GPT was trained on and memorized articles from The Times, with word-for-word copying in red and differences in black."
https://www.hollywoodreporter.com/business/business-news/court-advances-the-new-york-times-lawsuit-openai-1236173504/ https://www.hollywoodreporter.com/business/business-news/cou...
- 20k 5mo agoYes, LLMs fundamentally operate as a lossy compression scheme for their training data. There's been countless examples of them reproducing their training data with very high accuracy People claim that the data isn't stored, but clearly a representation of it is encoded and reproducible. I saw chatgpt word for word plagiarise a stack overflow comment just two days ago
- nonethewiser 5mo agoDoes this actually imply a representation of it has been stored or simply that the model is sort of over-fit?
- wat10000 5mo agoIs there a difference?
- nonethewiser 5mo agoWell yeah, if you're making the claim that it stores a representation of the data in some form. Does your calculator app store a representation of the answer to 1+2/2*1.1 and all other combinations of inputs or does it determine the answer from a set of rules?
- wat10000 5mo agoIt's a different case when the input contains more information than the output. If you put "1+2/2 x 1.1" into a calculator and it spit out a verbatim copy of a New York Times article, does it necessarily contain a representation, or does it just contain some really extensive rules? I'd argue those rules necessarily are a representation of that information, given that it contains far more information than provided by the input.
- scantron4 5mo agoAh but can it reply snarkily and close your ticket as a duplicate that is NOT A DUPLICATE? If not it will never recreate the real stack overflow experience.