Verification harness for quantum ML. A reproducible lab for stress-testing quantum models where predictive accuracy, identifiability, curvature, and robustness under noise can diverge.

Preregistered simulator and IBM hardware pilot for Lambda-IC gradient-scale diagnostics in qMERA trainability.

Companion notebook for A Technical Introduction to Quantum Neural Networks. Four small PennyLane experiments on encoding, depth and trainability, classical baselines, and finite-shot cost.

Curated GitHub Pages site tracking Quantum ML, Quantum NLP, Quantum Vision, and Hybrid Quantum-Classical AI - papers, architectures, LaTeX equations, circuit diagrams, and hardware milestones (2009–2026).

Simulations and analysis showing that gradient loss in noisy U(1)-equivariant quantum neural networks is governed by readout-visible sector coherence. Density-matrix simulations, regression analysis, and reproducibility code for a study of noise-induced gradient degradation in equivariant brickwork QNNs.

Quantum LeJEPA: applying LeJEPA self-supervised learning to quantum feature maps, and the second-moment operator M2 that predicts quantum-kernel trainability without training. Personal research.

A reproducible toolkit for auditing symmetry-organised complexity in equivariant quantum neural network ansatz, reporting sector occupation, cross-sector coherence, sectoral fluctuation, and generator-sum compliance against U(1), SU(2), and permutation symmetry before training.

Numerical code and source data for Barren Plateaus Beyond Observable Concentration

Systematic comparison of four VQC data-encoding strategies (Angle, Dense Angle, IQP, Data Re-uploading) for NLP sentiment analysis, with multi-seed validation on IMDb/SST-2 and IBM Quantum hardware verification.

Gradient-based analysis of barren plateaus in variational quantum classifiers with VQC, hybrid VQC, MLP baselines, and ansatz diagnostics

A representation-theoretic trajectory diagnostic for quantum neural networks. The symmetry-organised complexity index measures how a QNN distributes expressive capacity across the multiplicity structure of a target symmetry, rather than how much of Hilbert space it visits.

Weak inter-block coupling window in modular quantum circuits: entanglement/trainability tradeoff, depth law g*(L) ≈ 1.2/√L with a confirmed pre-registered prediction. Reproducible numpy POCs.