Current Projects
On the Efficacy of Solving the Poisson Equation through a QPINN
Vyaas
Physics-informed neural networks (PINNs) solve differential equations by training a neural net- work while penalizing violations of the governing equation and boundary conditions in the loss function [1]. Quantum physics-informed neural networks (QPINNs) extend this idea by embedding a parameterized quantum circuit within part of the classical network [2, 3]. This project will first test how three approaches solve the same one-dimensional Poisson prob- lem on $x∈[−1,1]$: a classical numerical method, a classical “vanilla” PINN, and a hybrid QPINN containing a quantum layer implemented with the Python library PennyLane. A fourth imple- mentation will discretize the equation and apply the Harrow–Hassidim–Lloyd (HHL) quantum linear-systems algorithm as a distinct quantum baseline [4]. The prescribed profile $u_0(x) = sin(πx)$ will be treated as a manufactured exact solution, since the time-independent Poisson equation has boundary conditions rather than an initial condition. After the 1D implementations are validated, the numerical, PINN, and QPINN pipelines will be extended to a 2D Poisson equation on $[−1,1]^2$; HHL will also be tested in 2D if the required circuit is worth it. We will measure relative L2 solution error against the exact solution, physics residual, convergence, training time, circuit resources, and variation across repeated runs. The study will alow the quantum methods to improve, match, or underperform the classical methods.
The project will isolate the effect of a quantum layer through controlled benchmarking rather than evaluating whether a hybrid model can produce a reasonable solution. Comparing the QPINN with a matched classical PINN, an exact solution, a conventional discretization, and HHL may clarify whether observed differences arise from representation, parameter count, optimization dynamics, discretization, quantum noise, or computational overhead.
Clear negative or neutral results are useful in quantum machine learning because they help identify where quantum components should be used and whether their use is justified. The benchmark, code, and analysis workflow developed here could also be reused by future QLab projects evaluating hybrid scientific-machine-learning models or quantum algorithms for computational fluid dynamics (CFD) and partial differential equations more generally.
Uncertainty-Aware Neural Decoding for Surface-Code Quantum Error Correction
Andrew, Lucas
Introduction: Quantum computers must detect and correct physical errors without directly measuring the encoded quantum state. Surface codes address this problem by repeatedly measuring local stabilizers on a two-dimensional qubit lattice. Changes in those measurements, called detection events, reveal where an error may have occurred, but they do not identify a unique physical-error pattern. A decoder must instead choose any correction that returns the state to the correct logical equivalence class. Minimum-weight perfect matching (MWPM) is a strong standard decoder, while recent neural decoders show that learned models can exploit spatial, temporal, and device-specific error correlations.
This project will test one narrow question: does preserving a neural model's uncertainty about physical errors improve surface-code decoding? A spatiotemporal neural network will map repeated syndrome measurements to per-qubit probabilities for the $X$ and $Z$ components of Pauli errors. A second, syndrome-consistent correction stage will receive those probabilities rather than a binary error map. We will compare this soft-information pipeline with MWPM, a direct neural decoder, and an otherwise matched model whose probabilities are thresholded before correction. The primary outcome will be logical error rate, not exact physical-error classification accuracy.
Intellectual Merit: The project will isolate the value of uncertainty through controlled ablation rather than evaluating only whether a neural decoder can correct errors. It will distinguish improvements caused by retaining probability information from improvements caused by model size, a learned syndrome representation, or a strong conventional decoding stage. The study will also compare physical-error inference metrics with the logically meaningful outcome, testing whether better per-qubit predictions actually produce fewer logical failures.
Broader Impact: Reliable decoding is necessary for useful fault-tolerant quantum computation, including future quantum-machine-learning workloads. A reproducible benchmark showing when soft neural information helps, has no effect, or hurts would help researchers avoid judging decoders by misleading physical-error accuracy alone. The data pipeline, ablation protocol, and logical-error analysis could also be reused by future QLab projects studying other codes, noise models, or quantum devices.
YBCO Superconductor with Intercalated Calcium
Huylam
Superconductors show promising applications in future electric grids in that they are capable of conducting electricity with zero resistance. Thus, they have the potential to massively increase the efficiency of current flow, which is applicable to commercial power grids and EV chargers. However, currently discovered superconductors exhibit superconducting properties only at low temperatures or extremely high pressures, with neither environment being practical enough to reproduce on a global scale. Furthermore, ceramic superconductors suffer from brittleness and poor heat dissipation, leading to breakdown caused by thermal stress. Therefore, more research is needed to find ways to address these shortcomings before superconductors can see widespread use. Yttrium-barium-copper-oxide (YBCO) shows desirable superconducting properties such as strong flux pinning, high critical current density, and a high critical temperature that allows cooling using liquid nitrogen. However, its natural brittleness can produce structural defects that weaken such properties. Therefore, this project will attempt to introduce calcium into the YBCO crystal lattice that will enhance its structural integrity while maintaining its superconducting properties.
Intellectual Merit The project will further explore how external materials can interact with YBCO crystals in ways that support superconducting. It can suggest standout materials for this purpose for further experimentation.
Broader Impact Research in materials science will contribute to overall knowledge needed to eventually create a room-temperature superconductor that can be used practically across different applications, from the power grid to computing. This would also reduce the need for cooling in circuitry.
2D Thermal Mapping of a Microchip Using Nitrogen-Vacancy Centers in Diamond
Owen, Maxwell
Introduction: Infrared sensors are currently the most popular method for industrial temperature mapping, but they have limitations when applied to complex materials and micrometer-level scales. For example, infrared wavelengths can range from 0.7μm to 20μm. In addition, factors such as the emissivity of a material, as well as changes in ambient temperature and humidity, drastically influence the accuracy of infrared sensors. Therefore, the aim of this project is to investigate an alternative option to thermal sensing that is more effective for mapping changes across microscopic electronic components. This project will utilize the nitrogen-vacancy (NV) center point defects inside a diamond plate to create a sensor capable of mapping thermal signatures across microchips on a micrometer scale. The goal of this project is to replicate the experimental procedure described by [1] and evaluate its accuracy by mapping heat signatures across resistors, capacitors, and transistors, gradually working up in complexity until finally mapping a microchip.
Intellectual Merit: The research advances the field of quantum thermal mapping by establishing a high-resolution, non-invasive wide-field imaging platform that uses the temperature-sensitive spins of NV centers in diamonds. By using optically detected magnetic resonance, our project will account for interference from separate magnetic-field effects, which have been major problems for NV thermo-mapping in the past. It also pushes the boundary of thermal sensing on a microscale, offering an alternative to traditional thermal mapping by providing highly precise localized temperature mapping without disrupting the electronics of the chip.
Broader Impact: The study helps the semiconductor and microelectronic industries by giving them a diagnostic tool to monitor thermal data in the new generation’s complex chips. By giving these industries a way of early detection of hot spots and other thermal stresses signifying structural issues, this imaging technology will help reduce error rates in microchip manufacturing and maintenance and improve the development of reliable electronic hardware. By integrating quantum mechanics, material science, and thermal engineering, it provides a foundation for future STEM professionals in the field of microelectronic technologies, which is a high priority for many nations around the world.
Sparse Uniformly Coupled Networks
Kevin
Introduction
A quantum network can be modeled as a graph, with vertices representing sites that can hold an excitation and edges representing couplings between those sites. An excitation may be spread across several sites in a superposition. As the state evolves, probability amplitudes combine through interference, changing where the excitation is likely to be found [1]. Adding a connection can therefore help or hurt transfer. Since each coupling also takes work to implement and calibrate, we have a reason to examine networks with relatively few edges.
We will investigate how connectivity affects the probability of transferring an excitation between two sites within a fixed time. Starting with minimally connected networks, we will compare the best transfer available when one or two additional edges are allowed. We will also measure what happens when a single edge is added to a particular network while the source–target distance stays the same. We will examine every small network in our stated range, then use a heuristic search to explore larger networks.
Intellectual Merit
Previous research has studied how to achieve reliable state transfer with limited network resources [2, 3]. Our project will compare transfer under four constraints, namely the number of sites, the number of extra edges, the source–target distance, and the available time. This comparison will help distinguish the benefit of generally higher edge density from the effect of placing an edge in a particular location. Our results will factor in improvements, decreases, and negligible differences in transfer across graphs.
Broader Impact
Our project aims to produce a database of small networks, with their transfer probabilities and the code used to calculate them. Future research groups could use these results to choose networks for further study or check their own simulations. Documenting the analysis and its limits will make our work easier to reproduce and build off of.
Quantum Input Selection for Cardiac MRI
Arnav
put introduction, intellectual merit, and broader impacts here
Photonic QKD
Ashwath, Ethan, Agastya, Isaac
put introduction, intellectual merit and broader impacts here
Holograph Data Storage
Avishi, Zoya
put introduction, intellectual merit, and broader impacts here
Rb MOT
Hridhaan, Jack, Sarah, Dev
A magneto-optical trap (MOT) cools atoms to microkelvin range temperatures by trapping atoms within perpendicular laser beams and varying magnetic fields [1]. A MOT uses opposing photon emission and absorption to reduce an atom’s speed, consequently reducing its temperature. By combining this with a magnetic field, the MOT works to trap an atom in a particular location to a near standstill (10 cm/s) [1]. At these extremely low temperatures, the atomic particles can be used to study phenomena such as Bose-Einstein condensation and enable various applications in quantum simulation, precision measurement, and quantum computing [2]. This project will aim to complete the MOT that the TJ Quantum Lab has worked on for the past two years. By constructing and optimizing a functional MOT system within the school environment, we hope to achieve ultracold atomic cooling of rubidium atoms. Once operational, this project will allow the TJ Quantum Lab to observe, study, and experiment with the unique quantum properties of ultracold atoms.
A MOT takes advantage of various quantum phenomena, including Doppler cooling and the Zeeman effect [1]. By building a MOT, we will attempt to manipulate atoms’ energy states and resonance to hold them at near zero velocity. If successful, we will be able to observe ultracold atoms’ quantum wave properties in Bose-Einstein condensate, a state of matter frequently researched for its macroscopic quantum behaviors.
By completing the MOT at TJ, we hope to provide a platform for future students to observe quantum behavior and produce new experiments that explore the fundamentals of quantum physics. Ultracold atoms are necessary for improving quantum computing and precision devices such as atomic clocks and stable qubit lattices [3]. With the cold neutral atoms the MOT produces, creating such devices becomes possible. Thus, we allow future students to pursue further understanding in quantum physics through research in the TJ Quantum Lab.
Electronic State Structure Near Cytosine Conical Intersections
Mikaeel, Justin
put introduction, intellectual merit, and broader impacts here
Rydberg atom reservoir learning
Elijah, Samuel
put introduction, intellectual merit and broader impacts here
Free-Space Model of Photonic Chip Meshes
Deepthi
put introduction, intellectual merit, and broader impacts here
Quantum Agents for Diabetic Retinopathy
Dhanvinkumar
Put introduction, intellectual merit, and broader impacts here
Rb Optically Pumped Magnetometry
Raina, Angela
put introduction, intellectual merit, and broader impacts here
NV-Diamond ODMR
Angelo, Anderson, Brian
put introduction, intellectual merit, and broader impacts here
Ketone Dynamics
Aarjav, Mohan
put introduction, intellectual merit, and broader impacts here
Quantum Prisoners Dilemma
Tuan, Shehani
Put Proposal Introduction, Intellectual Merit, and Broader Impacts here
Sparsification and Variable Reduction of Portfolio Optimization QUBOs with QAOA
Varun, Ethan, Ryan
Our project began with a simple question: if some covariances between stocks are very small, do we need to keep all of them when optimizing a portfolio? In a quantum circuit, retaining a pairwise interaction can add gates. Removing enough interactions might make the calculation easier, but it could also change which stocks are selected. We want to measure that tradeoff and investigate better ways to decide what can be removed. We will study a portfolio that selects a fixed number of stocks and gives them equal weights. Each stock has a binary variable indicating whether it is included. Expected returns, the covariance matrix, and a penalty for selecting the wrong number of stocks can be combined into a quadratic unconstrained binary optimization problem, or QUBO. QAOA, the Quantum Approximate Optimization Algorithm, will be used to search for solutions on small instances. Mr. Hannum suggested starting with a manageable mutual fund and considering whether reduction could go far enough to shrink the matrix itself. That adds a second question to the project. Removing covariance terms makes the matrix sparser, but it does not reduce the number of stock variables. To use fewer qubits in this encoding, we would need to fix or eliminate some of those variables. We will investigate both kinds of reduction and judge the resulting portfolios using the original objective.
We will study how sparsification and asset screening interact in fixed-size portfolio selection with QAOA. In particular, we will test whether their order changes the result: should we remove weak interactions before deciding which assets to exclude, or screen the assets using the full model first? Pruning can alter the portfolios used to rank assets, so the two orders may retain different stocks even when they produce problems of the same size. We will compare both orders with each reduction applied alone and measure the resulting portfolio loss and circuit cost. The proposed contribution is an empirical study of this interaction and its effect on QAOA. Both reductions have precedents, so we will compare the order experiment with the closest existing methods. Exact solutions on small cases and compiled circuit measurements will help distinguish information lost during preprocessing from limitations of QAOA.
Smaller circuits could make portfolio optimization experiments more practical with limited quantum resources. Our code and documented experiments could give later QLab groups a starting point. Identifying which assets or interactions cannot be removed without substantial loss would also help clarify when sparsification is useful.
QAOA Protein Groups
Nathan, Grace
Put proposal here
Connectivity in the Brain and Loss of Information
Isaac
Put abstract here