Tatum “Tate” Middleton

• Honors thesis: analyzing the semantic geometry of datasets through diffusion-model dynamics, using conditional mutual information phase transitions along denoising paths, continuing arXiv:2605.04830.
• Adapting classical MRF MAP inference into quantum adiabatic algorithms targeting quantum advantage on neutral-atom hardware; built an experimental topological Hopfield network model for Bose–Hubbard quantum computers.
• Simulating neutral-atom lattice dynamics with tensor-network methods (DMRG, TDVP via TeNPy).
• Explored stochastic path-integral (Martin–Siggia–Rose) methods to characterize the limits of computational advantage for ML on quantum and thermodynamic sampling platforms.
• Built early ML models for novel diffusion-model methods, precursor work to arXiv:2508.06614.

• Designed an optics experiment to test and verify attacks against BB84 quantum key distribution systems, validating attack models with the theory team using single-photon counters and FPGA tracking.
• Designed, simulated (Tidy3D), and tested photonic integrated circuits, including ring resonators and optical phased arrays.
• Built a 64-channel photonic hardware control system (PySide6/Qt) communicating with a Zynq FPGA for real-time control.

• Simulated and built a Fabry–Pérot optical cavity tuned to an integer-wavelength length for atomic control.
• Wrote a numerical simulation of Gaussian beam propagation through complex optics using ABCD matrices, accurate to 0.001 mm.
• Built a closed-loop PID system (Python, Raspberry Pi) correcting laser positional drift with 97% accuracy and millisecond correction times.

• Built a MATLAB image-processing pipeline that directly observed high-altitude gravity waves for the first time in NASA satellite imagery; co-authored the resulting paper.

Quantum Perceptron

A single-node quantum neural network in Qiskit that classifies MNIST digits with 85% accuracy.

M24 Quantum Cryptography Encoding Scheme

A quantum cryptography scheme based on quantum random walks for eavesdropping protection (PHYS 2170).

Gaussian Beam Propagation Simulator

ABCD-matrix simulation of Gaussian beams through multi-element optical systems, built for the Kaufman lab.

Powerwash

A full-stack web app for automatic data cleaning, including normalization and classification.

Graduate coursework: Quantum Information and Computing, Probability Theory.

Selected coursework: Quantum Mechanics 2, Thermodynamics & Statistical Mechanics, Electricity & Magnetism 2, Classical Mechanics & Mathematical Methods 2, Experimental Physics, Linear Algebra, Algorithms, Data Structures.

Graduate/postdoctoral summer school, accepted with full financial support. Topics included Hamiltonian learning and simulation, quantum error correction, and quantum hardware.

Formal methods with Yices and PVS, including applications to quantum compilers.