Otherwise, LLMs have most of the books memorised anyway: https://arstechnica.com/features/2025/06/study-metas-llama-3...
full transcript: pastebin.com/sMcVkuwd
By replacing the names with something unique, you'll get much more certainty.
So it might be there, by predcondiditioning latent space to the area of harry potter world, you make it so much more probable that the full spell list is regurgitated from online resources that were also read, while asking naive might get it sometimes, and sometimes not.
the books act like a hypnotic trigger, and may not represent a generalized skill. Hence why replacing with random words would help clarify. if you still get the origional spells, regurgitation confirmed, if it finds the spells, it could be doing what we think. An even better test would be to replace all spell references AND jumble chapters around. This way it cant even "know" where to "look" for the spell names from training.
I mean, you can try, but it won't be a definitive answer as to whether that knowledge truly exists or doesn't exist as it is encoded into the NN. It could take a lot of context from the books themselves to get to it.
Shoot, I'd even go so far as to write a script that takes in a bunch of text, reorganizes sentences, and outputs them in a random order with the secrets. Kind of like a "Where's Waldo?", but for text
Just a few casual thoughts.
I'm actually thinking about coming up with some interesting coding exercises that I can run across all models. I know we already have benchmarks, however some of the recent work I've done has really shown huge weak points in every model I've run them on.