Solomonoff Inductor
The Solomonoff inductor is a universal predictor centered on Solomonoff inductive inference. It assigns prior weights to all computable programs based on Kolmogorov complexity and produces a weighted mixture prediction over observed data. It is the foundation of general intelligence theories such as AIXI.
Timeline (last 90 days)
The Solomonoff inductor takes Solomonoff inductive inference as its core, assigning weights to all computable programs a priori based on program length, and then making weighted mixture predictions over observed data; it is the foundation of general intelligence theories such as AIXI.
If two Solomonoff inductors share the same training observations, they will give completely identical probabilistic predictions about the world, but the internal programs each uses to realize that prediction may have no similarity whatsoever.
All Facts (2)
The Solomonoff inductor takes Solomonoff inductive inference as its core, assigning weights to all computable programs a priori based on program length, and then making weighted mixture predictions over observed data; it is the foundation of general intelligence theories such as AIXI.
50%UnverifiedIf two Solomonoff inductors share the same training observations, they will give completely identical probabilistic predictions about the world, but the internal programs each uses to realize that prediction may have no similarity whatsoever.
50%