differentiable programming
A programming paradigm (called 'Software 2.0' by Karpathy) where programs are parameterized and optimized via gradient descent rather than hand-written; here applied to quantum and physics-based computers.
As defined in Guillaume Verdon
Where it comes up · 1
- Guillaume Verdon1 claim
Most important claims · 1 of 1Search all claims for “differentiable programming” →
- The universal meta-framework for programming physics-based computers is to parameterize the system so control parameters affect its physics, make those parameters differentiable, evaluate a loss function, and optimize via gradient descent—differentiable programming (Software 2.0) where gradient descent is a better programmer than you.Guillaume Verdon