Advancing the frontiers of quantum science and engineering

Abstract: I will summarize recent experimental progress at the Quantum Control Laboratory on using bosonic modes in the mechanical oscillators of a trapped ion for quantum computation, simulations, and sensing. First, I will discuss an optimal quantum control technique for preparing and manipulating highly non-classical bosonic states relevant to implementing hardware-efficient error-corrected quantum computing [1,2]. Our control strategy incorporates dynamical error suppression into numerically optimized temporal modulations of laser-driven spin-boson interactions. It enables coherent, deterministic, flexible, and error-robust preparations of a large variety of bosonic logical codewords, such as binomial codes and Gottesman-Kitaev-Preskill (GKP) codes [1]. Second, we implement a universal gate set for GKP logical qubits, including the first experimental realization of a two-qubit GKP entangling gate [2]. Our gate set is designed to avoid codeword distortions of physical GKP states. We create a GKP Bell state with a fidelity of 0.84, analyzed using a stabilizer sub-system decomposition formalism that allows efficient retrieval of GKP logical information. Furthermore, I will discuss a quantum sensing experiment that bypasses the precision restriction on incompatible observables (e.g., position and momentum) imposed by Heisenberg’s uncertainty principle. We use a single bosonic mode to measure modular observables and demonstrate uncertainties in position and momentum simultaneously reduced below the standard quantum limit [3]. If time permits, I will also discuss a “mixed qudit boson” approach to simulating quantum chemical dynamics, focusing on molecular problems that involve a strong vibronic (vibrational + electronic) coupling where the Born-Oppenheimer approximation breaks down [4,5,6]. References [1] Robust and deterministic preparation of bosonic logical states in a trapped ion, Phys. Rev. Lett. 133, 050602 (2024). [2] Universal quantum gate set for Gottesman-Kitaev-Preskill Logical Qubits, arXiv:2409.05455 (2024). [3] Quantum-enhanced multi-parameter sensing in a single mode, arXiv:2412.04865 (2024). [4] Analog quantum simulation of chemical dynamics, Chem. Sci. 12, 9794 (2021). [5] Predicting molecular vibronic spectra using time-domain analog quantum simulation, Chem. Sci. 14, 9439 (2023). [6] Direct observation of geometric-phase interference in dynamics around a conical intersection, Nat. Chem. 15, 1503 (2023).

Abstract: Modern AI tools significantly enhance quantum computing development and applications, particularly in quantum circuit optimization and quantum simulations. Moreover, Quantum Error Mitigation (QEM) aims to bridge the gap between the inaccuracies caused by noise and the accuracy required by useful quantum algorithms. We introduce a data-driven approach to mitigate the output distributions of current noisy quantum computers, without assumptions on the noise models. Our work studies several different AI models, from dense neural networks to transformers, that learn how to mitigate the noisy probability densities of quantum circuits by leveraging large datasets (144,000 to 225,000 circuits each): we design the models to exploit complex correlations between the circuit structure and the backend information and the models are agnostic to the circuit depth.

Venue: CQT Level 3 Seminar Room, S15-03-15 Title: Neutral atom array quantum processor at CQT

Abstract: Entanglement is a fundamental resource inquantum information processing, yet understanding its manipulation andtransformation remains a challenge. When it comes to manipulating entangledquantum systems on a single copy level, using entangled states as catalysts cansignificantly broaden the range of achievable transformations. Similar to theconcept of catalysis in chemistry, the entangled catalyst is returned unchangedat the end of the state manipulation procedure. In this talk we discuss recentprogress on entanglement catalysis, and also consider the more generalframework of entanglement batteries. An entanglement battery is an auxiliaryquantum system that facilitates quantum state transformations without a netloss of entanglement. We establish that reversible manipulation of entangledstates is achievable through local operations when augmented with anentanglement battery.Different entanglement quantifiers give rise to uniqueprinciples governing state transformations, effectively constituting diversemanifestations of a "second law" of entanglement manipulation.

Abstract: Complex systems are among the most challenging to study due to their intricate interplay of numerous components, encompassing the rich dynamics of many interacting particles, chemical reactions, biological processes, and pharmacological interactions. These systems reflect the complexity of life and the physical world, making their study vital across all branches of science.   Complex systems are challenging to model because they involve a vast number of interacting degrees of freedom. Capturing and predicting these interactions with high accuracy is daunting due to the sheer scale and complexity of behaviors and phases that can emerge. This diversity makes it difficult to create comprehensive models that can reliably predict system behavior across different scenarios.   Physicists often approach the challenge of complex systems through a technique known as coarse-graining. This method involves selectively simplifying a system by focusing on key variables while averaging out less critical internal degrees of freedom. By sacrificing some detail, this approach allows for greater predictability and understanding of the system’s behavior at larger length and timescales, as well as at higher levels of complexity.   Traditionally, coarse-graining has been an art, guided by the intuition and insight of experienced researchers. While this approach can yield powerful results, it often relies on strokes of genius and a fair share of luck. Now, machine learning is democratizing such coarse-graining through a process called tokenization. This tokenization equips researchers with the tools to identify optimal coarse-graining strategies efficiently and effectively. This technological advance empowers even the non-expert to uncover insights that were once accessible only to the most seasoned scientists.   In this talk, I will share the experience of how our students and postdocs, equipped with machine learning tools, have embarked on such journeys of data-driven discovery in complex systems. From uncovering novel spatiotemporal motifs in quantum materials, non-reciprocal interactions between biological cells, deciphering the language of functional disorder in piezoelectric materials, unveiling the structural origins of vibrant colors in butterfly wing scales, to creating novel computational lenses for cryo-electron microscopy. They have collectively pushed the boundaries of how we think about complex systems.   These experiences, I believe, show us how AI has become indispensable for understanding and discovering the secrets in complex systems.