As quantum computing moves toward error-corrected architectures, even familiar operations such as single-qubit rotations require new ways of thinking. In this hands-on session, participants will explore how operations can be implemented on logical qubits and why gate synthesis becomes a central challenge in fault-tolerant quantum computing. After a short introduction to Bloqade, the SDK tailored for researchers working with QuEra systems, participants will work through an activity that highlights the transition from gates acting directly on physical qubits to gates implemented through logical, error-corrected operations. The session will emphasize the practical tradeoffs involved in synthesizing rotations, including accuracy, circuit depth, and resource overhead.

Broadly applicable quantum advantage, particularly in classical data processing and machine learning, has been a fundamental open problem. In this work, we prove that a small quantum computer of polylogarithmic size can perform large-scale classification and dimension reduction on massive classical data by processing samples on the fly, whereas any classical machine achieving the same prediction performance requires exponentially larger size. Furthermore, classical machines that are exponentially larger yet below the required size need superpolynomially more samples and time. We validate these quantum advantages in real-world applications, including single-cell RNA sequencing and movie review sentiment analysis, demonstrating four to six orders of magnitude reduction in size with fewer than 60 logical qubits. These quantum advantages are enabled by quantum oracle sketching, an algorithm for accessing the classical world in quantum superposition using only random classical data samples. Combined with classical shadows, our algorithm circumvents the data loading and readout bottleneck to construct succinct classical models from massive classical data, a task provably impossible for any classical machine that is not exponentially larger than the quantum machine. These quantum advantages persist even when classical machines are granted unlimited time or if BPP=BQP, and rely only on the correctness of quantum mechanics. Together, our results establish machine learning on classical data as a broad and natural domain of quantum advantage and a fundamental test of quantum mechanics at the complexity frontier.

The discovery of quantum mechanics over a century ago revealed physical behaviors that don’t match our everyday intuition. Those ideas are now empowering the development of quantum computers to solve computational problems much faster than ever before. But how do we fit quantum ideas into the ways we currently think about computation, algorithms, and software? We will examine recent progress in quantum computers and quantum algorithms to highlight where quantum computing is already helping (or is likely to help) real-world applications. We will also explain the main obstacles to wider use, including error rates, qubit counts, and scaling up to larger systems. Most importantly, we focus on the new models of computation, new ways of designing algorithms, and new approaches to building and programming reliable quantum systems needed to unleash quantum acceleration.

Recent contributions are presented toward realizing fault-tolerant quantum computing at scale, including AI methods. Advances include more flexible choice of error correction codes, parallel methods of error decoding based on syndrome flips and GNN inference, cost modeling of logical transversals for distributed large-scale architectures, and novel simulation support for high-level operations with circuit-level error modeling.

Advances in large language models enable a new class of evolutionary workflows in which candidate quantum circuits are generated, evaluated, selected, and iteratively refined using automated feedback. We developed an LLM-guided evolutionary framework for finding compact quantum chemistry ansätze. It uses fermionic excitation operators, rather than using QASM gates. The operator pool includes paired double excitations, spin-adapted single excitations, opposite-spin double excitations, and same-spin double excitations. Operator selection and parameter initialization are guided by classical electronic structure information, obtained from selected CI calculations. The resulting ansatz is optimized and compressed by our LLM-guided evolutionary framework.

Symmetry plays a central role in quantum many-body physics, yet efficiently exploiting symmetries in quantum simulations remains an open challenge. In this talk, I will present a unified framework for incorporating symmetry group transforms into quantum algorithms through efficient symmetry-adapted projections onto irreducible representations of single or commuting groups. Our approach provides scalable quantum circuits and resource estimates for common symmetry groups, including cyclic and permutation groups. I will show applications of our work to condensed-matter and quantum chemistry simulations, including symmetry-adapted state preparation, quantum evolution, and phase estimation on noisy hardware. These results establish a practical framework for leveraging symmetries in digital quantum simulations and highlight their potential for achieving quantum advantage on fault-tolerant quantum computers.

Compiling many-body quantum dynamics into shallow, accurate, and hardware-efficient circuits is a central challenge in quantum simulation. In this talk, I will present a scalable approach to quantum compilation that combines tensor-network methods with ideas from quantum machine learning. Rather than relying solely on deterministic constructions such as Trotterization, these methods learn compact circuit representations of target dynamics from training data, enabling substantially lower gate costs while preserving high fidelity. These results show that tensor networks are not merely classical competitors to quantum computers, but powerful tools for training scalable quantum compilers. Across 1D, quasi-1D, and 2D settings, this framework achieves dramatic reductions in circuit depth, simulation error, and even fault-tolerant gate counts, pointing toward more practical quantum simulation on both near-term and future quantum hardware.

In this talk, we present a physically informed neural network (NN) framework for representing the effective interactions arising from coupled-cluster downfolding models for chemical systems and processes. This approach, termed the VNet model [1,2], enables efficient evaluation of effective interactions across a wide range of molecular geometries and correlation regimes, spanning varying levels of complexity in the underlying many-body wave functions. Beyond computational efficiency, the NN representation reveals a structured relationship between bare and effective interactions, which can be expressed through a tangent functional dependence on a set of latent variables. We refer to this characterization as the tangent model of the effective interaction. We further discuss how this tangent model connects to earlier theoretical analyses that quantify the differences between bare and effective Hamiltonians within corresponding active spaces. More broadly, VNet and related NN-based techniques provide a unifying framework that brings together diverse many-body methodologies, enabling the construction of effective Hamiltonians across distinct domains of quantum mechanics, including quantum chemistry and nuclear physics.

[1] S. Liang, K. Kowalski, C. Yang, N.P. Bauman, “Effective many-body interactions in reduced-dimensionality spaces through neural network models,” Phys. Rev. Res. 6, 043287 (2024).

[2] S. Liang, K. Kowalski, C. Yang, N.P. Bauman, “Exploring the nexus of many-body theories through neural network techniques: the tangent model,” Machine Learning: Science and Technology 6, 025040 (2025).

Artificial intelligence is opening new opportunities for quantum chemistry by accelerating the design, optimization, and interpretation of algorithms on both classical and quantum devices. In this talk, I will discuss how AI can be combined with reduced-density-matrix theory and quantum algorithms to simulate strongly correlated molecular systems more efficiently. Special emphasis will be placed on hybrid quantum-classical methods, including the contracted quantum eigensolver, where machine learning can help select operators, reduce circuit depth, and improve convergence. I will also discuss how these approaches connect classical electronic-structure theory with emerging quantum devices, creating new pathways for accurate molecular simulation in regimes where strong correlation, entanglement, and near degeneracy play central roles.

One of the main obstacles to the trainability of variational quantum algorithms and quantum machine learning models are barren plateaus, where the cost function (and its gradients) exponentially concentrates in parameter space as the size of the problem increases. We derive a formula for the variance of the cost function in terms of the dynamical Lie algebra (DLA) of the parametrized quantum circuit, i.e., the Lie algebra generated by the Hamiltonians in the circuit. We present a classification of DLAs generated by 1- and 2-local Pauli operators acting on a spin chain or more generally placed on the edges of an arbitrary interaction graph. Finally, we explicitly determine the DLA associated with the Quantum Approximate Optimization Algorithm with a Grover mixer. We prove that the dimension of the DLA grows polynomially with the number of qubits, and as a consequence, barren plateaus are avoided.

Significant machine learning progress has been realized in the past decade—from performing simple classification tasks to the advent of large language models. The accelerated adoption of deep neural networks can be partly accredited to the computational efficiency and accessibility of graphical processing units (GPUs). This success has resulted in an increased demand for compute-resource alternatives to GPUs. With the maturation of annealing quantum computing, research is being realized to place annealing quantum processing units (QPUs) as viable technology for machine learning applications. In this talk, I will present a practical approach to exploiting the sampling efficiency of annealing QPUs for performing neural network computations in discriminative modelling tasks. This demonstration consists of a series of experiments involving scalable models with varying degrees of contribution from annealing QPUs.

Generative models are powerful machine learning tools that can learn the probability density of a dataset and trained models have many uses in downstream tasks. The inherently probabilistic nature of these models makes them an interesting candidate for quantum algorithms. This talk will provide a brief overview of different constructions, training pipelines and data loading strategies to highlight current challenges and opportunities for development in this field.

Quantum will not replace classical computing—it will augment it as an accelerator inside real decision workflows. Yet most teams attempting quantum today hit the same walls: skills gaps, high experimentation costs, and uncertainty about where quantum helps (and where it doesn’t). In this keynote, we’ll share how SAS is approaching “applied quantum” by enabling side‑by‑side classical vs. quantum vs. hybrid comparisons for realistic use cases and by focusing on repeatability, defensibility, and cost efficiency. We’ll introduce the rationale behind SAS® Quantum Lab, a Viya-surfaced environment built to reduce the friction of experimentation through guided learning and development-time acceleration (parallelism, caching, and autotuning) before executing on QPUs.

Subspace methods are powerful, noise-resilient approaches for preparing ground states on quantum computers. The key challenge is obtaining a subspace with a small condition number that spans the states of interest while minimizing quantum resource requirements. In this talk, I will discuss how eigenvector continuation (EC) can be used to build such subspaces from the low-lying states of a set of Hamiltonians, with basis vectors prepared using truncated versions of standard state preparation methods — including imaginary time evolution (ITE), adiabatic state preparation (ASP), and the variational quantum eigensolver (VQE). By combining these truncated methods with EC, we directly improve upon them, achieving more accurate ground state energies at reduced cost, and demonstrate convergence even in challenging cases where ITE and ASP fail (e.g., ASP in the presence of level crossings and ITE with vanishing energy gaps).

In this session, participants will learn how to leverage PennyLane to design quantum algorithms. From inspiration to implementation, students will gain the practical software know-how needed to build the new frontier in QML research.

While variational algorithms have been widely studied in QML, the frontier of research is rapidly expanding into novel architectures. This hands-on session will step beyond the standard models and explore a new framework: train classical, deploy quantum.

Using PennyLane, an open-source platform for designing, compiling, and realizing meaningful quantum algorithms, participants will learn the core tools needed to construct and explore applications such as IQP circuit optimization. This session pushes beyond basic simulation and leverages PennyLane’s advanced ecosystem to explore generative machine learning tools, compilation at scale, and resource estimation.

Hardware platforms based on native continuous-variable (CV, oscillator) systems have attracted growing attention as an alternative to discrete-variable (DV, qubit) quantum systems. In this talk, I will highlight how hybrid CV-DV hardware offers a powerful computational paradigm by combining the complementary strengths of both CV and DV processors. I will present novel quantum control techniques and algorithms for CV-DV systems that enable new opportunities and applications in quantum error correction, quantum simulation, and quantum sensing. I will also highlight software tools for benchmarking and compiling these novel processors.

Recurrent Neural Networks (RNNs) are widely used to model complex temporal dependencies in sequential data. Quantum Recurrent Neural Networks (QRNNs) extend this idea to quantum machine learning by using quantum circuits inspired by classical recurrent architectures. In this paper, we propose three modifications to existing QRNN implementations that improve performance, generalization, and circuit efficiency. First, we evaluate the recently proposed EnQode approximate amplitude encoding subroutine, which offers the benefits of amplitude encoding while preserving shallow circuit depth. Second, we introduce a simple preprocessing strategy that augments amplitude encoded inputs with their pre-normalized magnitudes, improving generalization on two financial datasets. Third, we present a novel QRNN circuit architecture that is mathematically equivalent to the original model but significantly reduces circuit depth. We show that combining these three innovations should establish the new best practices for implementing a QRNN.

Efficiently representing and characterising quantum systems is challenging because Hilbert space is huge, but artificial intelligence, which is good at high-dimensional pattern recognition and function approximation, helps to predict quantum properties and to construct surrogates for quantum states. Applications include quantum certification and benchmarking to enhancing quantum algorithms and characterising strongly correlated phases of matter. arXiv:2509.04923

Quantum machine learning can be viewed not only as a search for speedups in classical machine tasks, but also as a way to learn from quantum data generated by quantum processors. In this talk, I will discuss this perspective through recent work on learning ground-state observables from quantum experiments. We use quantum data from approximate ground states of two-dimensional Heisenberg XXZ model, constructed using samples from IBM Heron quantum processors and classical high-performance computing, to train neural networks that predict observables across Hamiltonian parameter space. The results show accurate generalization to unseen parameters, suggesting a path toward using quantum computers as data generators for machine learning in many-body physics. 

Machine learning and artificial intelligence are often cited as application areas likely to benefit from quantum computing, but the timeline for realizing practical advantage remains an open question. Must meaningful quantum utility in machine learning wait for fault-tolerant hardware, or can it be explored in the current NISQ (Noisy Intermediate-Scale Quantum) era?
This talk introduces quantum computing and the state of quantum machine learning (QML) from an applied research perspective, with an emphasis on how quantum utility is being defined and evaluated. It highlights a promising near-term approach, quantum reservoir computing, and describes how it is being explored at SAS. The talk concludes with a demonstration of a QRC running in the new SAS Quantum Lab.

Trapped ions are a leading platform for quantum computation. In this tutorial, I will explain how single qubit gates are performed and how ion-motion operations can be used to either control the motional states of the ions or to enable two-qubit gates. I will briefly discuss applications of these controls to Hamiltonian simulation and quantum algorithms.

mRNA design is a structured biomedical optimization problem in which sequence choices influence secondary structure, stability, translation, and manufacturability. In this talk, I will discuss recent work on hybrid quantum-classical optimization for mRNA secondary-structure prediction and design, including dense-constraint QUBO formulations, compact quantum encodings, variational training, and problem-aware classical decoding.

A central example will be Pauli Correlation Encoding, which compresses many QUBO variables onto a smaller quantum register, together with a decoder that uses trained quantum expectation values as a prior for constrained solution search. Time permitting, I will connect these ideas to broader themes in quantum machine learning for biomedical applications, including data encoding, inductive bias, and learned quantum representations.

Electronic structure calculations on small systems such as H2, H2O, LiH, and BeH2 with chemical accuracy are still a challenge for the current generation of the noisy intermediate-scale quantum (NISQ) devices. One of the reasons is that due to the device limitations, only minimal basis sets are commonly applied in quantum chemical calculations, which allows one to keep the number of qubits employed in the calculations at minimum. However, the use of minimal basis sets leads to very large errors in the computed molecular energies as well as potential energy surface shapes. One way to increase the accuracy of electronic structure calculations is through the development of small basis sets better suited for quantum computing. In this work, we show that the use of adaptive basis sets, in which exponents and contraction coefficients depend on molecular structure, provide an easy way to dramatically improve the accuracy of quantum chemical calculations without the need to increase the basis set size and thus the number of qubits utilized in quantum circuits.