Expand description
Rust-native automatic-differentiation primitives and optional rule adapters. Rust-native differentiation primitives for QMBED scientific operations.
The native layer owns the mathematical derivative semantics. Optional
adapters such as the optional chainrules module only translate these tested primitives into
another protocol; they do not reimplement the derivatives.
Structs§
- Apply
Cotangents - State and parameter cotangents returned by an operator pullback.
- Apply
Jvp - Primal value and forward tangent from parameterized operator application.
- Apply
Pullback - One-shot pullback for
apply_vjp. - Gradient
Diagnostics - Backend-independent evidence returned with a native derivative.
- Ground
State Energy Gradient - Ground-state energy, Hellmann–Feynman gradient, and solver evidence.
- Parameter
Direction - Ordered forward perturbation of
ParameterValues. - Parameter
Gradient - Ordered reverse sensitivity of
ParameterValues. - Parameter
Schema - Stable names and differential domains for an ordered parameter vector.
- Parameter
Values - Ordered primal values for a parameterized operator.
Enums§
- Gradient
Status - Reliability classification attached to a scientific gradient.
- Parameter
Domain - Differential domain of one ordered operator parameter.
Functions§
- apply_
jvp - Evaluate a parameterized operator and its JVP without materializing
A(θ). - apply_
vjp - Evaluate a parameterized operator and prepare a one-shot reverse pullback.
- ground_
state_ energy_ gradient - Differentiate the isolated algebraic ground-state energy.