Factored Cognition
Decomposition of reasoning tasks
Last updated
Decomposition of reasoning tasks
Last updated
refers to the idea of breaking down (or factoring) sophisticated learning and reasoning tasks into many small and mostly independent tasks.
We’ll call programs that describe how to break down a task recipes.
We can think about machine learning systems on :
Process-based systems are built on human-understandable task decompositions, with direct supervision of reasoning steps.
Outcome-based systems are built on end-to-end optimization, with supervision of final results.
Factored cognition is a prerequisite for making the reasoning processes of large language models explicit so that they can be supervised more easily.