This is a great way to learn. Most people treat LLMs as black boxes, but building one from scratch — even tiny — forces you to understand attention, tokenization, and loss at a visceral level.
Curious what training data you used and how small you could go before the model stopped producing coherent output. There's an interesting cliff where models go from 'random tokens' to 'plausible text' and understanding where that threshold is teaches a lot about what these models actually learn.
How much training data did you end up needing for the fish personality to feel coherent? Curious what the minimum viable dataset looks like for something like this.
I love these kinds of educational implementations.
I want to really praise the (unintentional?) nod to Nagel, by limiting capabilities to representation of a fish, the user is immediately able to understand the constraints. It can only talk like a fish cause it’s very simple
Especially compared to public models, thats a really simple correspondence to grok intuitively (small LLM > only as verbose as a fish, larger LLM > more verbose) so kudos to the author for making that simple and fun.
> the user is immediately able to understand the constraints
Nagel's point was quite literally the opposite[1] of this, though. We can't understand what it must "be like to be a bat" because their mental model is so fundamentally different than ours. So using all the human language tokens in the world can't get us to truly understand what it's like to be a bat, or a guppy, or whatever. In fact, Nagel's point is arguably even stronger: there's no possible mental mapping between the experience of a bat and the experience of a human.
I’m not going to argue other than to say that you need to view the point from a third party perspective evaluating “fish” vs “more verbose thing,” such that the composition is the determinant of the complexity of interaction (which has a unique qualia per nagel)
Hence why it’s a “unintentional nod” not an instantiation
Curious what training data you used and how small you could go before the model stopped producing coherent output. There's an interesting cliff where models go from 'random tokens' to 'plausible text' and understanding where that threshold is teaches a lot about what these models actually learn.
Laughed loudly :-D
I want to really praise the (unintentional?) nod to Nagel, by limiting capabilities to representation of a fish, the user is immediately able to understand the constraints. It can only talk like a fish cause it’s very simple
Especially compared to public models, thats a really simple correspondence to grok intuitively (small LLM > only as verbose as a fish, larger LLM > more verbose) so kudos to the author for making that simple and fun.
Nagel's point was quite literally the opposite[1] of this, though. We can't understand what it must "be like to be a bat" because their mental model is so fundamentally different than ours. So using all the human language tokens in the world can't get us to truly understand what it's like to be a bat, or a guppy, or whatever. In fact, Nagel's point is arguably even stronger: there's no possible mental mapping between the experience of a bat and the experience of a human.
[1] https://www.sas.upenn.edu/~cavitch/pdf-library/Nagel_Bat.pdf
I’m not going to argue other than to say that you need to view the point from a third party perspective evaluating “fish” vs “more verbose thing,” such that the composition is the determinant of the complexity of interaction (which has a unique qualia per nagel)
Hence why it’s a “unintentional nod” not an instantiation