3 ms·
Good point. This gets us into the territory of not just "explainable" models, but also the ability to feed into those models "states" in a deterministic way. Th
by viksit 4y ago
Good point. This gets us into the territory of not just "explainable" models, but also the ability to feed into those models "states" in a deterministic way. This is a merger of statistical and symbolic methods in my mind -- and no way for us to achieve this today.
- sinenomine 4y agoWhy shouldn't we be able to just prompt for it, if our system models natural language well enough? ... And anyway, this problem of structured knowledge IO has been more or less solved recently: https://arxiv.org/abs/2110.07178 https://arxiv.org/abs/2110.07178