In briefDistributional variants of a quantum complexity conjecture
SummaryThis paper studies a conjecture asking whether quantum query algorithms under uniform random inputs can be simulated by classical ones with only polynomial overhead. The authors examine analogous conjectures for biased product distributions and for uniform distributions over Hamming-weight slices. They prove these variants are all equivalent to the original uniform-distribution conjecture. They also formulate distributional versions of the stronger Aaronson-Ambainis conjecture and show those are equivalent to the original as well.
In briefCircuit-level benchmarks of GKP-concatenated qLDPC codes: comparing finite-rate outer codes under realistic noise
SummaryThe paper benchmarks BB and tricycle quantum LDPC codes as outer codes for GKP concatenation using a unified simulation framework, from code-capacity noise to repeated noisy syndrome extraction with schedule-resolved displacement propagation. In the full circuit model, analog-informed BP-OSD decoding gives thresholds of about 0.212 for the selected BB sequence and 0.142 for the tricycle sequence, corresponding to squeezing requirements of roughly 10.46 dB and 13.94 dB. The results indicate analog GKP information still improves decoding under circuit noise, and the selected BB sequence shows a higher finite-size crossing estimate.
In briefGeneralized LIMDDs: Compact, Canonical Decision Diagrams Modulo a Group
SummaryThe paper studies decision diagrams for functions from {0,1}^n to C, where residual nodes are merged when they lie in one orbit of a group and the group element is recorded on edges. The group family is generated by phase rotations of order q with up to k control qubits, with or without bit flip X. The authors show exponential succinctness gains over Pauli-LIMDD, and give a five-rule reduction system with unique normal forms computable in polynomial time.
WhyGeneralizes Myhill-Nerode quotients to group actions with exponential succinctness gains, backed by proofs and separations; impact is confined to quantum decision diagrams.
Arend-Jan Quist, Alexis de Colnet, Thomas Reps et al.
In briefGauge freedom and efficient algorithms for learning open quantum system dynamics
SummaryThis paper studies learning a local Lindbladian, which describes open quantum system dynamics, in the presence of state-preparation-and-measurement (SPAM) noise. SPAM noise can make distinct Lindbladians experimentally indistinguishable, so part of the Lindbladian is unlearnable. The authors characterize this obstruction and present SPAM-robust algorithms that learn every universally gauge-invariant component using only trusted single-qubit operations.
WhyIntroduces gauge-aware characterization and SPAM-robust algorithms with sharp theoretical bounds, a substantive advance for open-system learning.
In briefAn exact polynomial bridge from task to prediction risk for time-multiplexed photonic quantum reservoirs
SummaryTime-multiplexed photonic reservoirs are usually judged by scores on a few datasets and designed by simulating candidate circuits one at a time. For time-unrolled passive linear networks with data-modulated gates or sources and a ridge readout, the authors derive an exact link from the task to the prediction risk in three steps: task features, encoding, and readout. They report that all exact statements agree with numerical checks on seven datasets to machine precision, so candidate architectures can be pruned without trial simulation.
In briefComputational complexity of Clifford template compilation and whether quantum computers help
SummaryThis paper studies how to choose repetition counts for a finite sequence of repeatable Clifford operations so that the resulting template realizes a target transformation of Pauli operators. For commuting templates with unrestricted integer exponents, feasibility is in NP∩BQP and a quantum algorithm can construct a solution. Restricting exponents to {0,1} makes feasibility NP-complete, while for commuting self-inverse cases binary feasibility is classically polynomial-time but bounding the total repetition count is NP-complete.
In briefA civilization technology scale combining the Kardashev and Barrow scales
SummaryThis paper proposes a technological capability scale that ranks civilizations by the number of control operations they can perform per second, where a control operation sets one binary degree of freedom of matter to a specified value. Drawing on constructor theory, it groups tasks into construction and its special cases on information media: processing, storage and transfer. The index Ξ = K + B adds the Kardashev index K to an efficiency index B based on energy per operation. The authors place humanity at K = 0.73 and B = -0.78, and describe an ultimate technology at Ξ = 7.58.
In briefExact distributions of complex spacing ratios in non-Hermitian random matrices
SummaryThis paper studies the complex spacing ratio, the ratio of complex displacements from a reference eigenvalue to its nearest and next-to-nearest neighbors, which diagnoses spectral correlations and nonintegrability in open quantum systems. Starting from exact joint eigenvalue densities, the authors derive finite-N distributions for Gaussian non-Hermitian random matrix ensembles in classes AI† and AII†. For AII† they give an exact algebraic expression for arbitrary N and evaluate distributions for N=3 to 6. For AI† they derive a normalized one-dimensional integral representation at N=3, analyze its asymptotics, and confirm the results by numerical diagonalization.
In briefNumerical Study of Laser Slowing, Cooling and Trapping of Positronium
SummaryThis paper numerically studies laser deceleration, cooling and trapping of fast ortho-positronium bunches produced in silica microchannel targets from positron bunches of a Surko trap. The authors report that standard Doppler cooling on the 1³S–2³P transition can simultaneously slow, cool and optically trap the atoms. Adding transverse lasers on the 2³P–3³D transition and including full Zeeman manifolds improves timing and final temperature. The required millijoule, roughly 100 ns laser pulses are said to be within current technology.
In briefRunning large IQP quantum circuit Born machines on smaller quantum hardware by injecting noise
SummaryQuantum circuit Born machines built from IQP circuits can be trained classically but need quantum computers for sampling, which limits them to hardware with enough qubits. The paper proposes a deployment protocol that injects noise at the logical level to split a large circuit into smaller clusters that fit on qubit-limited processors, with lightweight resampling of known noisy qubits to reduce the resulting error. Simulations and IBM Heron experiments deploying a 316-qubit circuit onto 156-qubit hardware show the mitigation brings the sampled distribution closer to the target. The method also preserves correlation structure more accurately than a classical baseline with accurate single-qubit marginals.
In briefWeakly interacting fermion Gibbs states are Gaussian mixtures and classically simulable
SummaryThe paper proves that for sufficiently weak interactions, the Gibbs state of a local fermionic Hamiltonian at any finite temperature and in any spatial dimension is a convex mixture of fermionic Gaussian states. Building on this structure, the authors construct a polynomial-time classical sampler that generates samples from the Gibbs state using Metropolis updates on a tree whose leaves are Gaussian states.
WhyRigorously proves weakly interacting fermionic Gibbs states are Gaussian mixtures, yielding a polynomial-time classical sampler.
In briefOverlap Gap Property: Showing Where Quantum Optimization Beats Quantum-Inspired Classical Methods
SummaryThe paper proves that the Mean-Field Approximate Optimization Algorithm, a classical quantum-inspired version of QAOA, cannot find arbitrarily near-optimal solutions for problems with the Overlap Gap Property. It links this limitation to finite-memory Approximate Message Passing algorithms. Numerical simulations on Max-4-XORSAT instances with this property show that QAOA can get past the barrier.
In briefQuantum codes with addressable transversal T gates that stay efficient at scale
SummaryThe paper constructs a family of binary CSS quantum codes with constant rate and linear distance. These codes support fully addressable transversal T gates, so any chosen tensor product of logical T powers is realized by physical T powers without later correction. The construction combines algebraic-geometry codes with a binary embedding, and the minimum embedding length is formulated as a minimum-weight problem, which improves the constants in the asymptotic bounds numerically.
In briefA method for approximating quantum correlation functions with accuracy and a rigorous error bound
SummaryThis paper addresses the classical simulation of non-integrable quantum many-body dynamics, which becomes intractable for larger systems. For the infinite-temperature correlation function, it derives a rigorous error bound for approximating the full dynamics within a computable finite-sized region. Applied to autocorrelations, the finite-region approach is shown to be especially accurate across one-dimensional systems and beyond. It can also provide rigorous error bounds for other approximate methods, demonstrated on matrix-product-operator simulations.
In briefDeriving the classical LLG equation from quantum spin dynamics
SummaryThe Landau-Lifshitz-Gilbert (LLG) equation describes dissipative magnetization dynamics. This paper shows that the LLG equation arises as the classical large-spin limit of the quantum LLG (QLLG) equation. The result is verified analytically and numerically for an interacting spin system. The authors also find that quantum effects become less noticeable as the number of spins in the Hamiltonian increases.
In briefQuantum state transfer for polynomial states via relative decoding
SummaryThe paper recasts decoded quantum interferometry (DQI) and its Hamiltonian extension (HDQI) as quantum state transduction through algebra homomorphisms, moving polynomial operator states from a source Hamiltonian to a target Hamiltonian. It proves that a relative distance greater than 2D is equivalent to preserving all inner products between degree-D polynomials. For a nonlinear pairwise optimization variant, the authors report an ideal quantum score of 0.643 versus 0.606 for the best tested classical heuristic.
In briefClassical limit on information rate for a dephasing-enhanced quantum neuron
SummaryThis paper studies a single-qubit dissipative quantum neuron updated by partial resets at rate ν and read out by repeated projective measurements. The authors derive an exact local Fisher information rate and show it is bounded by ν(1-s)(1+s)²/2 ≤ 16ν/27 near activation threshold. Optimized over readout interval and retained amplitude, the ideal bound cannot be exceeded by quantum coherence and is saturated by a classical two-state jump process.
In briefExperimental demonstration of Grover-based quantum-enhanced sensing
SummaryThis paper experimentally tests a theoretical proposal that repurposes Grover's search algorithm to better detect signals of unknown frequency across a wide detection bandwidth. The experiments used a single superconducting qubit coupled to a single superconducting cavity. Above a detection bandwidth of 10 MHz, the Grover-based protocol outperformed the non-Grover baseline, with the advantage growing superlinearly. For larger bandwidths and higher accuracy targets, it reduced the required signal by more than 10 times.
Mathieu Ouellet, Purnendu Sen, Xiangqin Wang et al.
In briefEnantiomer-specific pumping of chiral molecules in higher rotational states
SummaryThis paper proposes a scheme for enantiomer-specific state transfer of chiral molecules involving rotational states J=1 and J=2. It uses three linearly polarized microwave fields and one laser beam to pump one enantiomer into an enantiomer-specific dark state, while the other is dissipatively pumped out. The authors state the strategy can be extended to rotational states with higher angular momenta.
In briefSimulation-based inference of quantum system parameters with neural posterior estimation
SummaryThis paper proposes a likelihood-free framework that learns parameter posteriors for quantum systems directly from classical simulation data, addressing the intractable likelihoods caused by exponentially large Hilbert spaces. It pairs polynomial-cost classical simulators, such as Pauli propagation and tensor networks, with normalizing flows or other neural density estimators, so one trained model maps any new measurement record to its posterior in a single forward pass. The authors numerically demonstrate the approach on Pauli noise learning, quantum error mitigation, state tomography, and Hamiltonian learning, noting that posterior uncertainty also diagnoses non-identifiability.
In briefAll-or-nothing demonstration of preparation contextuality
SummaryThe paper introduces a parity-oblivious Hidden Matching task that demonstrates preparation contextuality in an all-or-nothing way. A quantum protocol using a single ⌈log₂n⌉-qubit message succeeds with certainty. The authors prove that no preparation-noncontextual model can achieve perfect success for any even n≥6.
WhyIntroduces a parity-oblivious hidden matching task giving all-vs-nothing preparation contextuality, with a hypercontractivity-based asymptotic bound and strong proofs.
Jayashree Karmakar, Rafiuddin Gazi, Biswadeep Chatterjee et al.
In briefFast classical simulation of noisy quantum memory by predicting error-free branches
SummaryThis paper studies how to classically simulate bucket-brigade quantum random access memory (QRAM) under noise. Noise histories are sampled, and branches whose routing paths avoid every sampled fault have predictable final states, so they can be pruned rather than evolved. The authors give a complete analysis for amplitude damping with qutrit and qubit encodings, and revise the pruning criterion for qubit encoding. The algorithm evolves only bad branches plus one reference branch, reaching speedups up to 285 times.
In briefDirectional pumping of a two-level system by a quantum source with fluctuation regulation
SummaryThe paper addresses why a resonant classical field only saturates a two-level system and cannot transfer population directionally. By keeping the excitation source as a quantum subsystem and tracing it out, the authors show that its second moments produce upward and downward dissipative channels. For a fully excited, number-diagonal source only the upward channel remains, giving monotonic pumping. With exponentially decaying source correlations, the pumping rate has a Lorentzian detuning dependence, and with thermal relaxation the steady-state excited population can exceed one half.
In briefA general measure of correlation for quantum and classical statistical systems
SummaryThis paper proposes a general measure of correlations for statistical systems called the correlation index. It applies to quantum and classical systems, in both equilibrium and nonequilibrium states. The authors define the index for arbitrary trace-class operators and illustrate it with examples of density operators and reduced density matrices. Equilibrium examples cover bosonic systems including Bose-Einstein condensates, superconducting systems, and spin systems. Nonequilibrium cases are considered for nonlinear dynamics of spin and pseudospin systems.
In briefGeneral Photon Subtraction from a Gaussian Perspective: a formalism describing photon-subtracted Gaussian states to improve cat, cubic phase, and GKP state generation.
SummaryThe paper introduces a formalism that applies a Gaussian blur and filter to the measured Fock Wigner function to describe general photon-subtracted Gaussian states. In this picture, initial squeezing, Gaussian measurements, and photon loss all contribute to the total blurring. The authors use it to trade off success rates, feed-forward compatibility, and loss tolerance when improving cat, cubic phase, and GKP state generation. They also introduce the expected Wigner logarithmic negativity as an additive monotone of non-Gaussianity and establish a general lower bound on resources for near-deterministic generation of target states.
Niklas Budinger, Ulrik L. Andersen, Peter van Loock
In briefSimulating Hybrid Qubit-Oscillator Quantum Circuits Using Gaussian Decompositions
SummaryThis paper presents a simulation framework for hybrid quantum systems combining discrete-variable and continuous-variable subsystems, based on linear combinations of Gaussian functions. The hybrid density matrix is split into blocks in a discrete basis, each represented as such an expansion, and analytic formulas are derived for Gaussian operations and controlled Gaussian channels with state-dependent noise. The approach avoids Fock-space truncation of the continuous-variable part. As an example, the authors simulate Gottesman–Kitaev–Preskill state generation in a cavity-QED system.
Kazufumi Tanji, Dot Belin Pio, Shohei Kiryu et al.
In briefCubic Phase Gate Using a Trapped Ion Oscillator
SummaryThis paper experimentally demonstrates a method to approximate the cubic phase gate, a nonlinear operation needed for universal continuous-variable quantum computation, using the motion of trapped ions. The authors numerically optimize the phases and amplitudes of state-dependent force pulses and verify performance for vacuum and low-amplitude coherent state inputs. The protocol is also said to extend to approximating other nonlinear unitaries.
Nigel Benjamin Lee Junsheng, Eugene Koh Jun Wei, Mu Young Kim et al.
In briefLog-Euclidean Rényi Conditional Mutual Information: A New Measure of Quantum Conditional Independence
SummaryThis paper introduces a quantum Rényi conditional mutual information that measures conditional independence between quantum systems. It is monotone under local quantum channels, additive under tensor products, and monotone in the Rényi parameter, converging to the quantum conditional mutual information as the parameter approaches 1. The authors derive bounds in terms of relative entropies to the Petz-recovered state, leading to improved upper and lower bounds on the QCMI and Rényi chain rules.
WhyProposes a quantum Rényi conditional mutual information satisfying all desired properties, with improved recovery-based bounds on QCMI.
In briefConstant-depth circuits for arbitrary single-qubit gate synthesis, with matching depth bounds
SummaryThe paper presents the first constant-depth circuit for synthesizing arbitrary single-qubit gates, a construction that is fully unitary and needs no pre-supplied catalyst. With only bounded-width gates, it achieves depth O(log log(1/ε)) and proves a matching lower bound. The authors also show that Hadamard, T, and generalized Toffoli gates suffice for constant-depth synthesis, and derive new structural results about shallow quantum circuit complexity.
WhyGives the first constant-depth circuit for arbitrary single-qubit synthesis and a matching lower bound, a substantial theoretical advance with complete proofs.
In briefThresholds and chaos in open-system digital quantum simulation
SummaryThis paper studies the Trotter step-size threshold in digital quantum simulation for open systems that include environmental noise and decoherence. It finds that beyond the threshold, simulation accuracy breaks down suddenly, and that this persists even in the steady-state limit. The authors use random matrix theory to characterise the onset of unitary and dissipative quantum chaos, and report that dissipative chaos emerges only when dissipation is strong enough but does not suppress the underlying chaotic structure.
Angsar Manatuly, Lukas M. Sieberer, Cahit Kargi et al.
In briefQuantum entanglement in quantum classifiers for breast cancer diagnosis
SummaryThis study compares three three-qubit variational quantum classifiers that differ only in how their qubits are entangled, testing each on the Wisconsin Diagnostic Breast Cancer dataset. The configuration with the highest entanglement gave slightly higher accuracy and F1-score and was the most stable across folds. The authors found the link between entanglement and performance weak and not significant, so it is an association rather than a proven cause.
In briefMicrosecond-programmable synthetic quantum matter for Heisenberg-limited metrology without adiabatic evolution
SummaryThis paper proposes a non-adiabatic Floquet protocol in a single-cavity Bose-Einstein condensate that uses rapid rotations and phase jumps to synthesize effective Two-Axis Twisting interactions from native longitudinal couplings. The authors simulate the dynamics with an exact matrix product operator encoding and DMRG/TDVP methods for N=10^3 ensembles, including decoherence. They report rapid generation of Heisenberg-limited spin squeezing, with a phenomenological optimal extraction time near 3.26 microseconds.
In briefFaster quantum algorithms for low-energy states of k-local Hamiltonians, with near-optimal density bounds
SummaryThis paper studies low-energy estimation and state preparation for general k-local Hamiltonians, two basic problems in quantum complexity theory. The authors present faster exponential quantum algorithms whose runtime exponent is governed by the binary entropy function. Their main technical result is an entropy-governed lower bound on the dimension of the Hamiltonian's low-energy subspace, obtained by depolarizing the ground state, which is optimal up to constant factors in the exponent for fixed k.
WhyAn entropy-governed lower bound on low-energy subspace dimension improves algorithmic exponents and is near-optimal for fixed k.
Sevag Gharibian, Fran\c{c}ois Le Gall, Ranitha Mataraarachchi et al.
In briefQuantum codes with separate recovery locality for bit-flip and phase-flip errors
SummaryThe paper introduces entanglement-assisted quantum locally recoverable codes with asymmetric locality, where bit-flip and phase-flip errors are corrected from recovery sets of different sizes. It extends a stabilizer-based sufficient condition and, in the CSS framework, places separate locality conditions on the two underlying classical codes. The authors derive Singleton-, Griesmer-, Plotkin-, sphere-packing-, and Cadambe–Mazumdar-type bounds for these codes. They also characterize when a pure CSS-type code built from two classical LRCs meets the Singleton-like bound with equality.
In briefEnergy gap of a confined chiral mode as a measure of critical zero-point fluctuations
SummaryThis paper studies the excitation gap where uniform and chiral superradiant boundaries meet in a quantum Rabi ring. The gap is proportional to the uniform zero-point variance divided by the frequency ratio η, closing as η^-2/3 while the variance grows as η^1/3. A quartic Schrödinger equation matches full spin-photon diagonalization, and a triangle of spins with Dzyaloshinskii-Moriya exchange follows the same relation with a different coefficient.
In briefDiscontinuity of Continuum Fermion Dynamics: how fermion time evolution can break continuity.
SummaryThe paper studies the Heisenberg dynamics of continuum fermions in R^d interacting through a pair potential. For a large class of nonconstant pair potentials, under suitable assumptions on the dispersion relation and external potential, the dynamics is shown not to be pointwise-norm continuous even on the gauge-invariant CAR algebra. An automatic-continuity argument then shows that neither the CAR algebra nor its gauge-invariant subalgebra is invariant under the dynamics, with non-invariance at almost all times.
In briefObservation of a topological defect state in the quantum Rabi model
SummaryThis paper realizes a topological defect state of the quantum Rabi model, a spin coupled to an oscillator, using a single trapped 171Yb+ ion. Reconstructed phase-space distributions show centroid trajectories with windings of one and zero in two regimes as the coupling phase varies, linking discrete-lattice topology to the oscillator's continuous phase space. Measured spin responses to drive-amplitude modulation relate the stability of spin polarization to energy gaps and allowed transitions.
In briefBound states in the continuum in cavity magnonics using Bragg and anti-Bragg mirrors
SummaryThis paper experimentally demonstrates bound states in the continuum (BICs) in cavity magnonics. Two periodic arrays of ferrimagnetic spheres act as mirrors in a dual-open-waveguide setup, forming Bragg and anti-Bragg cavities. A dark supermode couples to a probe magnon, producing a polariton splitting. Moving the probe sphere enables position-dependent detection of the cavity field.
In briefTopological edge state observed among localized bulk states in an anisotropic quantum Rabi model
SummaryThis study uses a trapped-ion quantum simulator to examine the topological edge state and bulk states of the anisotropic quantum Rabi model. Because the non-uniform couplings break translational symmetry, both the edge state and bulk states are localized, so spatial localization alone cannot distinguish them. The edge state shows well-defined chirality and near-complete spin-boson separability, which the bulk states do not. Phase-space tomography further shows its bosonic component is a squeezed vacuum state with squeezing up to 6.45 dB.
Sungjoo Lim, Chanyang Im, Christopher G. Yale et al.
In briefOptimal quantum algorithms for finding an item in a sorted list
SummaryThis paper studies ordered search, which finds a target in a sorted list of size n using comparison queries. It presents two new quantum algorithms, each using the optimal (1/π)ln n + o(log n) queries. The first is a zero-error algorithm derived from a continuum relaxation of the problem. The second is an exact algorithm based on an analytic solution of a polynomial program by Farhi, Goldstone, Gutmann, and Sipser.
In briefA spectral threshold guaranteeing non-maximal Schmidt number
SummaryThe paper asks when the eigenvalue spectrum of a bipartite state on C^d⊗C^d alone ensures its Schmidt number is not maximal. Using spectral bounds for (d-1)-block positive operators, it proves that states whose largest eigenvalue does not exceed the sum of their p smallest eigenvalues, for p ≤ ⌊d²/2⌋, have Schmidt number at most d-1 under all global unitaries. It also shows that states satisfying the reduction criterion under arbitrary global unitaries belong to this class, while for d ≥ 3 it constructs maximal-Schmidt-number states satisfying the reduction criterion on both subsystems, showing the global unitary requirement is essential.
In briefRényi channel capacity does not rise with non-signaling assistance
SummaryThe paper gives an alternative variational proof that shared non-signaling correlations between sender and receiver cannot increase the Rényi capacity of a classical channel. The result follows from a channel-divergence data-processing inequality by Oufkir, Yao and Berta, and uses a Legendre transform to write capacity as an extremum over affine functionals. Combined with separate arguments for Shannon capacity and endpoint orders, it establishes monotonicity for all orders. The authors also derive consequences for random-coding, sphere-packing and strong-converse exponents.
Tina Shariat, Matt Hoogsteder-Riera, Marco Tomamichel
In briefQuantum memory in an effective Λ-type system using dynamic detuning control of a superconducting qubit pair
SummaryThis theoretical paper studies storing quantum light in an effective Λ-type system made of two tunable superconducting qubits coupled to a semi-infinite transmission line and spaced half a wavelength apart. Their frequency detuning couples a superradiant excited state to a subradiant metastable state. By dynamically adjusting this detuning, an incoming quantum field is mapped into the subradiant collective state, stored as a long-lived excitation, and retrieved on demand. The authors propose the system as a minimal, tunable quantum memory for scalable superconducting quantum networks.
In briefNakayama's reduction of quantum topos and Bayesian quantum computing
SummaryThis paper presents Nakayama's method for limiting contexts in Döring-Isham topos quantum theory, unifying this limitation with the definition of probability as a single topology choice. The authors explain how this unification supports a Bayesian view in which probability changes as data is obtained. They describe this change as a local shift in the applicability of a predictive theory and apply it to quantum computing.
In briefEquivalence of topological indices for U(1)⋊Z₂-symmetric quantum spin chains
SummaryThis paper studies symmetry-protected topological phases in integer-spin quantum chains with on-site U(1)⋊Z₂ symmetry. It proves that Ogata's operator-algebraic index and Tasaki's elementary twist index are equivalent, giving an explicit error bound for computing the index. Combined with a boundary-field gap theorem and earlier calculations, it shows that the infinite S=1 antiferromagnetic Heisenberg chain has a symmetric gapped ground state with both indices equal to −1.
WhyProves equivalence of two topological indices with explicit error bound for U(1)⋊Z2 spin chains, a substantive but narrowly scoped mathematical physics result.
In briefTensor-engineered van der Waals NbOCl2 metasurface for generating polarization-entangled photon pairs
SummaryThis paper presents a resonant metasurface built from NbOCl2, a van der Waals nonlinear material, that uses its optical anisotropy to control polarization-dependent nonlinear responses. By tuning resonant modes, the nonlinear response along the c axis is enhanced by three orders of magnitude, and the second-order susceptibility tensor is redistributed between orthogonal polarization channels. The approach yields polarization-entangled photon pairs with a measured fidelity of up to 92%.
In briefNative CP2K implementation of the machine-learned Skala exchange-correlation functional for molecular and solid-state calculations
SummaryThe paper implements the machine-learned Skala exchange-correlation functional natively in CP2K's GPW and GAPW methods, using a joint one-centre reconstruction of density, gradients, and kinetic-energy density. For crystalline CO2, NH3, and urea, all-electron Skala-D3(BJ) calculations give a mean absolute error of 1.54 kJ/mol against diffusion Monte Carlo. For the 13 DMC-ICE13 phases, the absolute lattice-energy MAE is 1.16 kJ/mol. However, the LC10 benchmark shows systematically underestimated equilibrium lattice constants, indicating structural overbinding.
Johann Pototschnig, Franz P\"oschel, J\"urg Hutter et al.
In briefQBX: A Compiler for Simulating 2-local Qubit Hamiltonians on Quantum Chiplets
SummaryThis paper presents QBX, the first quantum compiler built for simulating 2-local qubit Hamiltonians on quantum chiplet architectures. It uses a scalable hierarchical approach and a highway mechanism to reduce the number and distance of cross-chiplet communications. According to the authors' evaluations, QBX outperforms both general-purpose and domain-specific compilers in circuit depth and operation count, and scales to larger simulations.
In briefMoving Quantum Annealing Errors into a Cavity
SummaryQuantum annealing slows down at small energy gaps, and the errors made there go undetected. The authors propose cavity quantum annealing, which couples a high-frequency cavity mode mid-schedule to swap these errors into photons. In the p-spin model, annealing time drops to nearly the square root of the time needed without the cavity. Counting photons also certifies results, with high-photon runs reaching above 99.99% ground-state fidelity much sooner.
In briefWhy approximating quantum code distance is computationally hard
SummaryThis paper studies how hard it is to approximate the minimum distance of a stabilizer quantum code. Earlier reductions only reached an additive gap of order √N, and this work closes that gap to a linear one. It also begins a fine-grained analysis, giving running-time lower bounds tied to SETH and Gap-ETH.
WhyCloses the sqrt(N) gap to linear additive hardness for quantum code distance and gives first fine-grained lower bounds, with rigorous proofs.
In briefQuantum-computer Kohn-Sham DFT using orthonormalized plane-wave orbitals and a dual-basis eigensolver
SummaryThe paper presents an all-electron, full-potential Kohn-Sham density functional theory method for quantum computers. It uses a quantum-circuit eigensolver to solve self-consistent field problems in a plane-wave-orthonormalized-orbital basis, solving each step in a contracted basis. For an H8 chain, emulated results agree with classical calculations within 0.02 mHa. On IonQ Forte-1 hardware, the measured energy difference matches the noise-free value within one standard deviation.
In briefControlling two-photon interference with structured pump light to tune the shift from constructive to destructive effects
SummaryTwo photons entering a Hong-Ou-Mandel interferometer bunch or anti-bunch depending on their state's symmetry, producing constructive or destructive interference. This paper tailors the pump field in spontaneous parametric down-conversion, using rotating structured pump modes in Hermite-Gaussian and Laguerre-Gaussian bases. The authors report that the pump's parity shapes the transition from constructive to destructive interference, with sensitivity increasing with mode order in the Hermite-Gaussian case. For Laguerre-Gaussian modes, angular resolution improves by an integer factor proportional to orbital angular momentum content, suggesting potential for Hong-Ou-Mandel-enhanced rotation sensing.
In briefDynamically Controlling Photon Statistics in a Raman-Coupled Cavity QED System
SummaryThe paper proposes a scheme that actively switches photon statistics between bunching and antibunching through quantum interference. It uses a Raman-coupled cavity-atom system with parametric amplification, where time-modulating the intermediate-state detuning either deepens photon blockade or enhances bunching. Simultaneously modulating the cavity and intermediate-state detunings produces antibunching unattainable under static conditions, and it recurs periodically at intervals set by the modulation frequency.
In briefQuantum data processing equality: a paper turning the quantum data processing inequality into an exact equality
SummaryThe paper promotes the quantum data processing inequality, which states that quantum relative entropy does not increase under quantum channels, to an exact equality using boundary and surface Poisson integral representations across a complex strip. The relative-entropy loss splits into two nonnegative components: a Petz recovery term governing state reconstructibility and a boundary term measuring mismatch of relative modular dynamics. The authors also establish recovery bounds at any prescribed modular parameter, with optimal square-root scaling in the relative entropy loss.
WhyPromotes quantum data processing inequality to an exact equality via integral representations, decomposing loss into Petz recovery and boundary terms.
In briefAutonomous phase discovery: a paper on automatically exploring quantum phases with differentiable programming and unsupervised learning
SummaryThe paper introduces a fully autonomous system that combines differentiable programming and unsupervised learning to discover quantum phases of matter without phase labels or a prescribed target phase. The search evaluates ground-state data along an adaptive trajectory rather than a fixed parameter grid. On a generalized cluster chain with up to 200 distinct phases, it finds up to 25 more phases at the same budget and matches random sampling given thirty times its budget. On a 50-parameter Chern insulator, it reaches sectors that the simple harmonic constructions considered do not obtain.
WhyCombines differentiable programming with unsupervised learning for adaptive phase discovery, benchmarked against random sampling on two models with clear gains.
In briefStudy on how embedding design affects quantum physics-informed neural networks
SummaryThis paper studies how the embedding that maps coordinates into quantum states affects quantum physics-informed neural networks (QPINNs) that solve PDEs. The authors introduce a unified embedding framework and propose two embeddings, LQNN-TE-QPINN and AdaFreq-QPINN. They evaluate them on one- and two-dimensional Burgers equations against analytical encodings and other architectures. The results indicate that embedding design substantially influences approximation and optimization behavior.
Ban Q. Tran, Nahid Binandeh Dehaghani, Susan Mengel et al.
In briefA quantum linear-system algorithm optimal for both oracle query counts
SummaryThis paper asks whether one quantum algorithm for linear systems can meet the known optimal query bounds for both the matrix block-encoding oracle and the right-hand-side state-preparation oracle at once. Given a constant-factor estimate of the solution norm, the authors present an algorithm using O(κ log(1/ε)) matrix queries and O(κ/s) state-preparation queries. They also prove matching lower bounds on a single Hermitian instance, showing the two bounds are jointly optimal.
WhyResolves an open question on simultaneous query optimality, with a matching same-instance lower bound and rigorous proofs.
