3 ms·
All these weird partial language thought patterns im surprised none of the teams have taught the models to think in something like court stenography or some ver
by cchance 2mo ago
All these weird partial language thought patterns im surprised none of the teams have taught the models to think in something like court stenography or some very dense pattern (i mean they even tried caveman language at one point)
- chaboud 2mo agoKeep in mind that the model is thinking in a token space, itself a compressive representation of language. (Note: there's still a huge grammar penalty, so, ugh do think small.)
- kzrdude 2mo agoIt selects tokens but they expand to embedding vectors which are huge, also in memory and attention requirements, I think?
- qeternity 2mo agoThe real breakthrough is going to be thinking in latent space.
- rcxdude 2mo agoArguably this is already happening: the whole state of the model gets fed through from token to token, and even just shoving a bunch of dashes in between the input tokens and the model's output can improve performance (thinking tokens from the model help a little bit more, but the difference is not as large as you mught expect).
- robkop 2mo agoOne of the main goals at the moment is to keep thinking human legible. You can imagine how much harder it’d be to do root cause analysis on the recent OpenAI event if we couldn’t even tell what they’re thinking.
- rcxdude 2mo agoIt's not obvious if that would help. Some tests have shown that what exactly the thinking tokens are only makes a small difference to the performance of the model, and that the contents of them are sometimes only tenuously related to to what the model actually does after them. It seems like it could be they are more like a kind of "mumbling" and that the underlying mechanism by which they actually help performance is just that it makes more computation available to the model by just giving more passes on the earlier input tokens through the network.