Watch a point move toward (0,0)
one move: track partial derivatives
Open the code pane immediately, then use the Spec tab, optional design question, tests, and artifact when you want them.
Finish each move before carrying the artifact forward.
For several variables the gradient is a vector of partial derivatives.
Deliverable: Write the mechanism in one sentence.Minimize f(x,y)=x²+y² by updating both coordinates.
Deliverable: Type the smallest runnable implementation.Both coordinates and the function value must converge together.
Deliverable: Record the expected output and one edge case.Use unequal starting points and learning rates; observe axis-wise motion.
Deliverable: Change one variable and explain the result.A two-parameter optimizer with a convergence report.
Deliverable: Save the code, output, and a short failure note.Both coordinates and the function value must converge together.
Use unequal starting points and learning rates; observe axis-wise motion.
A two-parameter optimizer with a convergence report.
Mapped from the supplied Build Everything PDF, source page 20. The five moves are Deriva’s implementation contract for this project.