Re:Learning Universal Predictors
by cubefox on Tuesday January 30, 2024 @06:34PM
(#39197527)
Interesting that they combine assumptions from highly idealized abstractions (universal Turing machines, Solomonoff induction) with realistic neural network architectures.
[ Parent
| on HN ]
They have been doing this forever. Here is a paper by some of the same authors: A Monte Carlo AIXI Approximation (2009). https://arxiv.org/abs/0909.0801