Episode 26: Sugandha Sharma, MIT, on biologically inspired neural architectures, how memories can be implemented, and control theory
Generally Intelligent
00:00
Using SBD, You Can Represent Non-Parametric Distributions
SBD allows you to represent non-parametric distributions. The final model generalizes the other types of tasks, like waffle tasks. So it's sort of trained on lots of different distribution. And then again, you do the standard Bayesian inference, but in this lower-dimensional space.
Transcript
Play full episode
