Landscape Tools is a Python library designed to visualize, characterize, and analyze the loss landscapes of variational quantum algorithms.
It provides utilities for:
- One-dimensional, two-dimensional, and three-dimensional loss scans
- PCA-based landscape projections
- Optimization trajectory visualization
- Interpretable parameter-sensitivity analysis directly on quantum circuits
- Exploratory landscape analysis
- Barren plateau analysis through variance-based scaling studies over circuit depth, number of qubits, observables, and Pauli-string padding strategies
Built with quantum machine learning experiments in mind, Landscape Tools helps researchers diagnose trainability issues, inspect optimization dynamics, and better understand how variational quantum ansätze behave across parameter space.
pip install landscape_toolsfrom landscape_tools import landscape_visualization, barren_plateaus, exploratory_landscape_analysisMore details are available in the documentation.
 Gabriel Legros |
 Henri de Boutray |