February 2024

A key challenge in realizing fault-tolerant quantum computers is circuit optimization. Focusing on the most expensive gates in fault-tolerant quantum computation (namely, the T gates), we address the problem of T-count optimization, i.e., minimizing the number of T gates that are needed to implement a given circuit. To achieve this, we develop AlphaTensor-Quantum, a method based on deep reinforcement learning that exploits the relationship between optimizing T-count and tensor decomposition. Unlike existing methods for T-count optimization, AlphaTensor-Quantum can incorporate domain-specific knowledge about quantum computation and leverage gadgets, which significantly reduces the T-count of the optimized circuits. AlphaTensor-Quantum outperforms the existing methods for T-count optimization on a set of arithmetic benchmarks (even when compared without making use of gadgets). Remarkably, it discovers an efficient algorithm akin to Karatsuba’s method for multiplication in finite fields. AlphaTensor-Quantum also finds the best human-designed solutions for relevant arithmetic computations used in Shor’s algorithm and for quantum chemistry simulation, thus demonstrating it can save hundreds of hours of research by optimizing relevant quantum circuits in a fully automated way.

We study homogenous and isotropic quantum cosmology using the spinfoam formalism of Loop Quantum Gravity (LQG). We define a coupling of a scalar field to the 4-dimensional Lorentzian Engle-Pereira-Rovelli-Livine (EPRL) spinfoam model. We employ the numerical method of complex critical points to investigate the model on two different simplicial complexes: the triangulations of a single hypercube and two connected hypercubes. We find nontrivial implications for the effective cosmological dynamics. In the single-hypercube model, the numerical results suggest an effective Friedmann equation with a scalar density that contains higher-order derivatives and a scalar potential. The scalar potential plays a role similar to a positive cosmological constant and drives an accelerated expansion of the universe. The double-hypercubes model resembles a symmetric cosmic bounce, and a similar effective Friedmann equation emerges with higher-order derivative terms in the effective scalar density, whereas the scalar potential becomes negligible.

Quantum gravitational tunneling effects are expected to give rise to a number of interesting observable phenomena, including, in particular, the evolution of black holes at the end of their existence or the emergence of the early universe from a quantum phase. Covariant Loop Quantum Gravity provides a framework to study these phenomena, yet a precise identification of tunneling processes is still not known. Motivated by this question, we consider a related, simpler case, that of Ponzano-Regge amplitudes: we find a surprising and detailed analogy of a class of simple transition amplitudes with tunneling processes in non-relativistic quantum mechanics.

Quantum Cellular Automaton (QCA) is a model for universal quantum computation and a natural candidate for digital quantum simulation of relativistic quantum fields. Here we introduce the first photonic platform for implementing QCA-simulation of a free relativistic Dirac quantum field in 1+1 dimension, through a Dirac Quantum Cellular Automaton (DQCA). Encoding the field position degree of freedom in the Orbital Angular Momentum (OAM) of single photons, our state-of-the-art setup experimentally realizes 8 steps of a DQCA, with the possibility of having complete control over the input OAM state preparation and the output measurement making use of two spatial light modulators. Therefore, studying the distribution in the OAM space at each step, we were able to reproduce the time evolution of the free Dirac field observing, the Zitterbewegung, an oscillatory movement extremely difficult to see in real case experimental scenario that is a signature of the interference of particle and antiparticle states. The accordance between the expected and measured Zitterbewegung oscillations certifies the simulator performances, paving the way towards the application of photonic platforms to the simulation of more complex relativistic effects.

The study of quantum reference frames (QRFs) is motivated by the idea of taking into account the quantum properties of the reference frames that we use, explicitly or implicitly, in our description of physical systems. Like a classical reference frame, a QRF can be used to define physical quantities such as time, position, momentum, and spin relationally. Unlike its classical analogue, it relativises the notions of superposition and entanglement. Here, we provide a novel explanation for the frame-dependence of superposition and entanglement by tracing it back to the question of how configurations or locations are identified across different branches in superposition. We show that, in the presence of symmetries, whether a system is in ‘the same’ or ‘different’ configurations across the branches depends on the choice of QRF. Thus, sameness and difference-and, as a result, superposition and entanglement-lose their absolute meaning. We apply these ideas to semi-classical spacetimes in superposition and use coincidences of four scalar fields to construct a comparison map between the spacetime points in the different branches. This allows us to determine whether a given event is located at ‘the same’ or ‘different’ points in the superposed spacetimes. Since this feature depends on the choice of QRF, we argue that the localisation of an event should not be seen as an inherent property. This alleviates previously voiced concerns that QRF changes could have empirical consequences for interference experiments, such as the BMV proposal. Moreover, it implies that the number of events is equal in both the flat and the curved spacetime implementations of indefinite causal order. We conclude with the ‘quantum hole argument’ as a generalisation of Einstein’s hole argument, arguing that not just spacetime points but also their identification across a superposition lose their absolute physical meaning.

Recent progress in table-top experiments offers the opportunity to show for the first time that gravity is not compatible with a classical description. In all current experimental proposals, such as the generation of gravitationally induced entanglement between two quantum sources of gravity, gravitational effects can be explained with the Newton potential, namely in a regime that is consistent with the weak-field limit of general relativity and does not probe the field nature of gravity. Hence, the Newtonian origin of the effects is a limitation to the conclusions on the nature of gravity that can be drawn from these experiments. Here, we identify two effects that overcome this limitation: they cannot be reproduced using the Newton potential and are independent of graviton emission. First, we show that the interaction between a generic quantum source of gravity, e.g. in a wide Gaussian state, and a test particle cannot be reproduced with the Newton potential nor with a known classical theory or gravity. Hence, observing the form of this interaction would require either a modification to classical gravity or its quantum description. Second, we show that the quantum commutator between the gravitational field and its canonically conjugate momentum appears as an additional term in the relative phase of a generic quantum source interacting with a test particle. Observing this term in the phase would be a test of the gravitational field as a quantum mediator. Identifying stronger quantum aspects of gravity than those reproducible with the Newton potential is crucial to prove the nonclassicality of the gravitational field and to plan a new generation of experiments testing quantum aspects of gravity in a broader sense than what proposed so far.

In this paper, we provide a comprehensive study of asymptotically flat spacetime in even dimensions $dgeq 4$. We analyze the most general boundary condition and asymptotic symmetry compatible with Penrose’s definition of asymptotic null infinity $mathscr{I}$ through conformal compactification. Following Penrose’s prescription and using a minimal version of the Bondi-Sachs gauge, we show that $mathscr{I}$ is naturally equipped with a Carrollian stress tensor whose radial derivative defines the asymptotic Weyl tensor. This analysis describes asymptotic infinity as a stretched horizon in the conformally compactified spacetime. We establish that charge aspects conservation can be written as Carrollian Bianchi identities for the asymptotic Weyl tensor. We then provide a covariant renormalization for the asymptotic symplectic potential, which results in a finite symplectic flux and asymptotic charges.