Isn’t TLP’s email on his website?
metachirality
Karma: 438
[Question] Why do we enjoy music?
Sure, I just prefer a native bookmarking function.
I wish I could bookmark comments/shortform posts.
You can actually use this to do the sleeping beauty experiment IRL and thereby test SIA vs SSA. Unfortunately you can only get results if you’re the one being put under.
This sort of begs the question of why we don’t observe other companies assassinating whistleblowers.
I think there should be a way to find the highest rated shortform posts.
I like to phrase it as “the path to simplicity involves a lot of detours.” Yes, Newtonian mechanics doesn’t account for the orbit of Mercury but it turned out there was an even simpler, more parsimonious theory, general relativity, waiting for us.
Vanessa Kosoy has a list specifically for her alignment agenda but is probably applicable to agent foundations in general: https://www.alignmentforum.org/posts/fsGEyCYhqs7AWwdCe/learning-theoretic-agenda-reading-list
We don’t actually know if it’s GPT 4.5 for sure. It could be an alternative training run that preceded the current version of ChatGPT 4 or even a different model entirely.
I think it disambiguates by saying it’s specifically a crux as in “double crux”
Copied from a reply on lukehmiles’ short form:
The hypothesis I would immediately come up with is that less traditionally masculine AMAB people are inclined towards less physical pursuits.
If it is related to IQ, however, this is less plausible, although perhaps some sort of selection effect is happening here.
The hypothesis I would immediately come up with is that less traditionally masculine AMAB people are inclined towards less physical pursuits.
This feels like Scott Alexander could’ve written something about, and it has the same revelatory quality.
I assume OP thought that there was some specific place in the training data the LLM was replicating.
I think that requires labeled data.
It doesn’t and the developers don’t label the data. The LLM learns that these categories exist during training because they can and it helps minimize the loss function.
I don’t think there are necessarily any specific examples in the training data. LLMs can generalize to text outside of the training distribution.
Another problem is, why should we expect to be in the particles rather than just in the wave function directly? Both MWI and Bohmian mechanics have the wave function, after all. It might be the case that there are particles bouncing around but the branch of the wave function we live in has no relation to the positions of the particles.
Have you tried just copying and pasting an alignment research paper (or other materials) into a base model (or sufficiently base model-like modes of a model) to see how it completes it?
I don’t think I’m really looking for something like that, since it doesn’t touch on the perception of music as much as it does the reasons why we have it.