Quantum Mind#
Quantum robotics: computing, sensing and cognition, each against its classical baseline.
Quantum Mind is an open-source Python library for the three threads of quantum robotics: quantum computing for robot algorithms (robot problems as optimisation, with classical baselines and a Qiskit QAOA adapter), quantum sensing (quantum-limited sensor models and their fusion with classical sensors) and quantum-probability cognition (how people judge, decide and trust, and how a robot should read them). Every model sits next to the classical model it has to beat, and every belief ends in a cost-aware decision.
pip install "quantum-mind[all]"
Multi-robot allocation, grasp selection and slot scheduling as QUBOs, classical baselines and a thin Qiskit QAOA adapter.
Quantum limits, atom-interferometer and MEMS accelerometer models, and a Kalman filter that fuses them.
COBRA, a context-rotated density-matrix belief filter, with quantum-like human models and the robot decision layer.
Question order, quantum Bayesian networks, trust dynamics and bistable perception, each with classical baselines. Four families.
Ask or act, question planning, calibration and conformal sets, fusion, trust-aware hand-over and allocation, ROS 2.
Gate-by-gate trajectories, animations, interactive HTML, live widgets and tomography on real back ends.
What is new in this library#
Classical filters treat evidence as arriving in neutral, independent pieces. People and robots do not: a look changes what a hand movement means, a question changes the answer to the next one, and asking about trust changes trust. The quantum-probability formalism represents exactly this, and Quantum Mind turns it into tested code:
COBRA, a belief filter whose generalised measurement is rotated by the previous frame, with the Bayes filter and an incoherent filter as built-in limits so the effect of context and the effect of coherence can be separated. On a synthetic task with a context effect it lowers decision cost by about 0.02 against a Bayes filter given enough data (significant in 2 of 5 seeds), but it is worse with little data and slightly worse where context does not matter; the benchmark, the ablations and the cases where it loses are reported in full.
Trust as an open quantum system, an order-aware hand-over policy and a value-of-information question planner whose answers depend on the order of the questions.
A decision layer that never reports a belief without a cost, a calibration check and a classical baseline.
It runs on ordinary computers and claims no quantum computational advantage.
See the qubits#
A Qiskit circuit, gate by gate. While the two qubits are entangled their Bloch vectors move inside the sphere. Drag to rotate; press play or use the slider.
from qiskit import QuantumCircuit
from quantum_mind.viz import circuit_trajectory, animate_bloch, save_html
qc = QuantumCircuit(2); qc.h(0); qc.cx(0, 1); qc.ry(0.6, 1)
save_html(animate_bloch(circuit_trajectory(qc, steps=15)), 'bell.html')
The two-qubit playground runs the same simulation in the browser.
Where to go next#
Install, then fit, compare and simulate in a few lines.
Quantum-like cognition from the basics: projections, order effects, interference.
Sixteen step-by-step tutorials, from the introduction to the three threads and a robot loop, with real outputs and figures.
A block diagram for every model: inputs, internals, outputs, baselines and references.
The equations of every model, as the code computes them, from the Born rule to COBRA.
The layers, every model family, circuits, hardware and the viewer.
20 runnable scripts on quantum-like models, COBRA, robotics, perception and the viewer.
Interactive Bloch spheres, robot questioning and trust.
Every public class and function, with parameters, returns and examples.
Which model do I need, glossary, FAQ and what is validated.
Changelog, citing, licence and contributing.