Skip to content

API Reference

Generated directly from the package's own docstrings via mkdocstrings — nothing here is duplicated by hand, so it can't drift out of sync with the code. Scoped to the dense_evolution package itself; dashboard_core, ia_utils, and mcp_server (the Composer kernel and its MCP adapter) are built on top of these modules but aren't part of this generated reference — see below.

Driving these modules without writing Python: every module in the table has a real, tested execution path through the Composer kernel, reachable either from a browser (Composer) or from an MCP-aware agent (MCP Server, 22 tools — Claude Code, Claude Desktop, ...). Both call the exact same kernel, not a separate reimplementation.

Module What it's for
Simulator Core dense statevector engine (DenseSVSimulator)
MPS Simulator Matrix-product-state engine for low-entanglement circuits at scale
Chunk Anti-OOM statevector chunking, including distributed multi-device dispatch
QASM Parser OpenQASM 2.0 / 3.0 parsing
Compiler Circuit transpilation (QuantumTranspiler)
Registry & Noise Models Hardware detection, NoiseModel Kraus channels
Gates Gate matrix tables (GATES, PARAMETRIC_GATES, GATE_IDS)
Mitigation Zero-Noise Extrapolation, scalar and density-matrix
Healing Predictive-healing primitives
Interop Qiskit / PennyLane bridges
Autodiff Differentiable circuit-to-energy pipeline for VQE
Harrison Tight-Binding Universal (materials-independent) sp3 tight-binding Hamiltonians
VHD Tight-Binding Material-specific sp3s* tight-binding, validated against real GaAs/Si/Ge gaps
Fermions Majorana-fermion → qubit (Jordan-Wigner) mapping
Entropy Multi-qubit partial trace, von Neumann entropy, mutual information
Trotter Real-time Hamiltonian evolution as an actual gate circuit
Native Hartree-Fock From-scratch JAX/Obara-Saika ab-initio HF engine for elements outside PennyLane's own STO-3G table
QEC Erasure-aware stabilizer decoding — Pauli commutation, syndromes, and location-aware error correction