If even some hypothesis “very close” to the current hypotheses + priors were missing for in-context learning, then you’d get a) or b). If all hypotheses close to the current hypothesis + priors could be explored with near-full Bayesian accuracy, but there was some limit, some metric under which which things “further away” in that metric space both took more evidence to reach and also had more and more of the possible hypotheses simply missing and not creatable during in-context learning, then you’d get c).
There’s a limit in how far I want to go brainstorming capabilities improvements, but basically what I was suggesting is that an obvious approach one might try is first learning things in-context, then doing some form of SGD imitation learning from that to train a model that now already knows how to do that and doesn’t need to use a lot of context to figure it out.
If even some hypothesis “very close” to the current hypotheses + priors were missing for in-context learning, then you’d get a) or b). If all hypotheses close to the current hypothesis + priors could be explored with near-full Bayesian accuracy, but there was some limit, some metric under which which things “further away” in that metric space both took more evidence to reach and also had more and more of the possible hypotheses simply missing and not creatable during in-context learning, then you’d get c).
There’s a limit in how far I want to go brainstorming capabilities improvements, but basically what I was suggesting is that an obvious approach one might try is first learning things in-context, then doing some form of SGD imitation learning from that to train a model that now already knows how to do that and doesn’t need to use a lot of context to figure it out.