Theoretical
Free
Validate algorithm structure and theoretical bounds before any cost is incurred. Perfect gates, infinite coherence.
- Qubits
- Unlimited (ideal quantum computer)
- Audience
- Research papers, theoretical bounds, algorithm design
00Quantum machine-learning research
QMANN
Quantum memory, classical learning.
QMANN studies how classical learning can be combined with quantum memory operations. Its reference path separates Theoretical, Simulation and Hardware modes, with a cost estimate before hardware work. Read each result in the context of its mode and baseline.
- Reproducible modes
- 3
- Qubits in simulation
- 20
- Qubits on hardware
- 4–12
- Open licence
- Apache-2.0
01Execution modes
Pick the mode, not the marketing.
Most QML libraries hide the cost surface and the noise model. QMANN exposes both, and ships three explicit modes so you always know what your run actually means.
A simulated result is not a hardware result; the hardware route alone does not establish a general learning advantage.
Simulation
Free (compute only)
Classical simulation with noise modelling and quantum-inspired operations. Ship the science before paying for shots.
- Qubits
- Up to 20 (with noise modelling)
- Audience
- Algorithm validation, development, education, reproducible research
Hardware
Paid (estimate before submission)
Run on real quantum devices with cost-estimator gating. Hardware mode requires explicit credentials and a confirmed budget.
- Qubits
- 4–12 (NISQ-bounded)
- Audience
- Bounded hardware experiments with an approved budget
Compare modes before moving to hardware.
Theory can shape a question; simulation can test a defined setup. A hardware run needs its own cost, environment and result record before anyone compares it with a classical baseline.
3 modes
20-qubit simulation
Hardware budget gate
- 01
Frame
Define the learning question, classical baseline, target metric, and acceptable compute budget.
Input · question + baseline
- 02
Bound
Use theoretical mode to test algorithm structure and expected limits without claiming hardware behaviour.
Artifact · analytical bounds
- 03
Simulate
Run the controlled noise model and compare the result with the classical baseline before spending shots.
Evidence · reproducible run
- 04
Approve
Review qubit count, provider route, credentials, estimated shot cost, and the reason hardware is necessary.
Gate · explicit budget
- 05
Record
Compile accepted hardware work through qmesh and retain the signed run context, result, and limitations.
Output · provenance record
Four components, hybrid by design.
Model core
Q-Matrix
Quantum memory layer with entangled qubit registers.
Quantum-LSTM
Hybrid controller with parameterised quantum circuits.
Evaluation and use
Energy-optimal protocols
Advanced error mitigation and measurement strategies.
Real-world applications
Healthcare, industrial IoT, autonomous systems modules.
Citations, not vibes.
Every technique below maps to a published reference. Pull the paper, read the section, then run the code.
01Quantum LSTM with segment processingEPJ Quantum Technology · March 2025
QSegRNN-style segment-based quantum recurrent neural networks. Quantum gates for LSTM operations, hybrid memory cells, enhanced temporal pattern recognition.
02QAOA with warm-start adaptive biasPhysical Review · 2025; EPJ Quantum Technology · August 2025
Conditional diffusion-based parameter generation. Classical solution warm-start, adaptive bias correction. Faster convergence with reduced optimisation overhead.
03Grover dynamics optimisationCornell Lawler Research · January 2025
Grover-inspired amplitude amplification for optimisation problems. O(√N) speedup for unstructured search with adaptive oracle construction.
04Quantum-enhanced transformersarXiv:2504.00068 · arXiv:2501.15630
Quantum-classical attention integration. Multi-head quantum attention, entanglement-based correlations, configurable quantum attention ratio (0–100%).
05Circuit-noise-resilient virtual distillationCommunications Physics · October 2024
Enhanced virtual distillation with noise resilience. Multiple virtual copies, adaptive thresholds, maintains effectiveness with imperfect gates.
06Learning-based error mitigationRecent QEM literature · 2025
ML models for quantum error prediction and correction. Neural network error models, adaptive strategies, real-time error prediction and correction.
These are reference themes to inspect, not evidence of measured QMANN performance.
A1Application and stack annexSector pipelines and qmesh provenance path
Sector — example pipelines
- Finance
- QAOA portfolio optimisation, fraud detection on transaction streams, market prediction with quantum LSTMs, quantum-enhanced risk modelling and stress testing.
- Drug discovery
- Quantum transformers for molecular property prediction, drug-target binding affinity, Grover-dynamics molecule generation, ADMET safety prediction.
- Materials science
- Quantum property prediction, adaptive-VQE crystal structure refinement, battery material design, catalyst design via quantum simulation.
- Healthcare
- Pattern recognition for diagnosis, optimisation for personalised medicine, protein folding via quantum simulation, clinical-trial patient matching.
- Industrial
- Quantum anomaly detection for predictive maintenance, classifiers for defect detection, QAOA for logistics, quantum simulation for chemical processes.
QMANN compiles through qmesh.
When you target hardware, QMANN circuits flow through qmesh for IR lowering, error mitigation, and ed25519-signed run manifests. The same provenance chain that satisfies regulated industries also gives ML researchers fully reproducible benchmarks.
06Quick start
Pick a mode. Run an example.
qmann / examplesSH
$ git clone https://github.com/neuraparse/QMANN.git $ cd QMANN $ pip install -e . # 🔬 Theoretical mode (FREE) $ python examples/01_theoretical_mode.py # 💻 Simulation mode (FREE) $ python examples/02_simulation_mode.py # ⚛️ Hardware mode (PAID — estimate first!) $ python scripts/estimate_hardware_costs.py --qubits 6 --shots 1000 $ python examples/03_hardware_mode.py
Open research