System One Decision Model
A class of decision models that treat the input as a complete state and output a structured probability distribution through a single forward inference pass. It supports three question types—choice, noul, and score—and is suited to automated workflows that require reliable judgments.
Timeline (last 90 days)
The System One decision model is a model that does not generate answers token by token; it reads state and a structured question, returns the probability of each option in a single inference, and its output is a structured probability distribution rather than free text.
The protocol defines three question types: choice selects from named options and returns confidence and distribution; noul judges whether a proposition is true or false and returns P(true); score locates a position on an ordered scale and returns the expected level and distribution.
All Facts (2)
The System One decision model is a model that does not generate answers token by token; it reads state and a structured question, returns the probability of each option in a single inference, and its output is a structured probability distribution rather than free text.
50%UnverifiedThe protocol defines three question types: choice selects from named options and returns confidence and distribution; noul judges whether a proposition is true or false and returns P(true); score locates a position on an ordered scale and returns the expected level and distribution.
50%