In briefQuantum routing and perfect state transfer on signed graphs with noise
SummaryThe paper proposes a deterministic quantum routing architecture based on coined Szegedy quantum walks on edge-duplicated signed graphs, treating edge signs as local phase shifts. This enforces zero back-scattering at routing nodes. The authors demonstrate perfect state transfer with unit fidelity at exact arrival times on structures such as the signed dumbbell switch and glued binary trees. They also analyse routing performance under amplitude and phase damping noise.
In briefFault-tolerant quantum phases: a classification of mixed-state phases by their ability to protect and transfer quantum information
SummaryThe paper classifies mixed quantum states into phases based on whether they can preserve and fault-tolerantly transfer logical quantum information. It shows that a noise recovery threshold is a property of an entire phase, and applies this to subsystem codes, demonstrating phase equivalence between 2D and 3D color codes. The authors also study how syndrome structure affects fault tolerance in circuits with measurements.
WhyProposes a fault-tolerant classification of mixed-state phases, proving thresholds are phase-wide properties with rigorous theoretical results.
In briefParallel algebraic surgery for qLDPC codes with constant shuttling depth
SummaryThis paper studies high-rate quantum LDPC codes on neutral-atom arrays and introduces algebraic surgery for lifted product codes, co-designing logical measurements, auxiliary codes, and atom shuttling. The protocols enable hardware-efficient Pauli-product measurements, including parallel logical measurements, with constant shuttling depth under stated layout assumptions. In a representative hardware-model benchmark, a cycle supporting 33 parallel canonical logical measurements is about 2x faster than compared single-observable graph-based surgery schedules, with competitive qubit overhead.
In briefLowering Entanglement Cost of Distributed Bipartite Quantum Computation With Few Extra Qubits
SummaryThis paper studies distributed quantum computing, where several quantum processors use quantum communication to jointly run large computations, with entanglement consumption measuring communication cost. For exact deterministic implementations of bipartite unitaries, it shows that the operator Schmidt rank lower bound on entanglement cost can be reached for every bipartite Clifford unitary using at most two auxiliary qubits per processor. For non-Clifford unitaries given by exact Clifford+T decompositions with T-count t, the constructed implementations use the same qubit overhead and need at most 2t Bell pairs above that lower bound.
WhyShows the operator Schmidt rank entanglement lower bound is attainable for bipartite Clifford unitaries with two ancillas per QPU, plus a bound for non-Clifford cases.
In briefReversing quantum state diffusion using differential geometry
SummaryQuantum state diffusion turns a Lindblad master equation into random trajectories of pure quantum states, described by a probability density on the manifold of pure states. The paper derives the time reversal of this diffusion, obtaining a backward stochastic equation for the state vector valid for any Hilbert space dimension and any Lindbladian. This equation is used to generate a new ensemble of trajectories close to the original. For the depolarizing channel, the authors bound the error of the generated distribution compared with the original.
In briefOptical measurement is nearly quantum: universal optical limits inspired by quantum physics
SummaryThe authors develop a formalism for optical sensing, inspired by quantum information theory, that abstracts away specific optical systems and derives general limits on measurement based only on the quantity being measured. They report a universal bound governing any passive optical system regardless of its technology, which is attained by a unique measurement scheme for each quantity of interest. The optical formalism is said to be nearly identical to the quantum one, including an accompanying uncertainty principle.
In briefWhether EPR pair count can predict runtime in distributed quantum computing
SummaryThis paper examines whether EPR-pair consumption, a common communication-cost objective in distributed quantum computing, reliably reflects execution time. The authors compare EPR cost and runtime across all balanced two-QPU mappings of QFT, QAOA, and CDKM circuits. They find that mappings with equal EPR cost can differ substantially in runtime, and lower-EPR mappings can be slower. The reliability depends on the workload and communication conditions, so EPR count and execution time are related but distinct objectives.
In briefA quantum anonymous voting protocol that lets voters pick multiple candidates and tallies votes itself
SummaryThis paper proposes a self-tallying quantum anonymous voting protocol that supports multiple-selection voting, which prior protocols of this type could not do. Voters receive particles from entangled states and encode their ballots into measurement-derived vectors. Each voter can verify their own ballot, and any participant can compute the vote count for each candidate. The authors state the protocol satisfies self-tallying, nonreusability, verifiability, and fairness.
In briefAll-Microwave Multiqubit Gates: single-step gates for superconducting qubits without tunable couplers
SummaryThe paper proposes an all-microwave scheme for single-step multiqubit gates with n control and m target qubits on frequency-tunable superconducting transmon networks. It uses cross-resonance drives to create effective ZX interactions without tunable couplers, with the effective Hamiltonian derived analytically. Numerical simulations with pulse optimization show three-qubit gates exceeding 99.9% fidelity in 60 ns, and the approach is extended to a five-qubit CXXXX gate.
In briefDistributed quantum sensing of dynamic radio-frequency fields using time-bin entangled qubits
SummaryThis paper proposes a discrete-variable distributed quantum sensing framework that replaces polarization probes with time-bin entangled qubits. The temporal structure of time-bin Bell states serves as a tunable built-in differential sampling reference for sensing radio-frequency fields. When the time-bin separation matches the RF half-period, the authors report a theoretical 6 dB sensitivity gain over one-sample static polarization encoding, validated by Monte Carlo simulations.
Vedansh Nehra, Richard Birrittella, Benjamin Malia et al.
In briefImproved quantum random self-reduction for linear problems
SummaryThis paper studies quantum random self-reductions for linear problems over finite fields. Given an oracle that matches a linear map only on a fraction of inputs, the authors give a uniform quantum reduction that computes the matrix-vector product on any chosen input with probability at least 2/3. The running time improves on a previous reduction, and the method avoids explicitly learning a subspace that was computationally costly before.
WhyImproves quantum random self-reduction runtime for linear problems from O~(n^{3/2}+T) to O~(nT^{1/3}) by avoiding explicit subspace learning, with rigorous theoretical analysis.
Vahid R. Asadi, Shuichi Hirahara, Nobutaka Shimizu
In briefLearning random quantum circuits and when pseudorandomness emerges
SummaryThe paper presents an efficient algorithm that learns k-dimensional brickwork random quantum circuits using only copies of the output state obtained by applying the circuit U to the all-zero input. For depth-d circuits on n sites with random 2ℓ-qubit gates, it learns the circuit with high probability in poly(n, 2^{ℓd}) time on any constant-dimensional lattice. The algorithm runs in polynomial time whenever ℓd = O(log n).
WhyProposes a local correlation criterion avoiding backward light-cone learning and a dimension-independent anticoncentration inequality, a clear improvement with substantial technical support.
In briefSharp limits on how fast quantum entropy can mix
SummaryThe paper studies how fast the von Neumann entropy of a quantum ensemble can change under Hamiltonian evolution of its components. Using an exact integral representation of the commutator obtained from the operator layer cake theorem of Cheng and Liu, it proves the optimal constant c=1 conjectured by Mariën, Audenaert, Van Acoleyen, and Verstraete. The result gives a dimension-independent sharp bound on mixing rates and implies an entangling-rate bound for bipartite Hamiltonians.
WhyProves the sharp constant c=1 for Bravyi's conjecture via an operator layer cake representation, also correcting a Lieb-Vershynina conjecture on general ensembles.
In briefDouble localization for spectral gaps of quantum Gibbs samplers
SummaryThis paper introduces double localization, a framework that proves spectral gaps for quantum Gibbs samplers by combining geometric localization with interface localization. The authors apply it to Hamiltonians built from the vertex terms of Kitaev's finite-group quantum double model, without plaquette terms, on finite cubic graphs of girth at least six. For fixed nontrivial finite groups and a range of inverse temperatures, they prove a spectral-gap lower bound that does not depend on graph size. For the group S3, the bound holds uniformly over the whole interval 0 ≤ βJ ≤ log 3.[The original text says
Ryu Hayakawa, Angus Southwell, Caesnan M. G. Leditto et al.
In briefImpossibility theorems for dual Hamiltonians of non-Abelian lattice gauge theories
SummaryThis paper studies Hamiltonian simulation of lattice gauge theories, relevant to understanding QCD. It proves that for non-Abelian theories, reparametrising the gauge configuration while preserving the number of degrees of freedom and fundamental commutation relations inevitably leads to long-range interactions in the electric part of the Hamiltonian. The authors try to mitigate this by adding gauge links and constraints, which works for U(1) but is shown to be inconsistent for non-Abelian lattice gauge theories.
WhyProves nonlocality is unavoidable for non-Abelian dual Hamiltonians under degree-preserving reparametrisation, and that the extended construction is inconsistent.
Marco Garofalo, Tobias Hartung, Timo Jakobs et al.
In briefQuantum circuits using measurements and feedforward built from error-correcting codes
SummaryThe paper shows that measurements and feedforward let shallow quantum circuits deterministically implement global unitaries and prepare long-range entangled states. It establishes a general correspondence between these protocols and quantum error-correcting codes, where unitary feedforward removes post-selection exactly when the measurement projectors are detectable errors on the codespace. Using non-Pauli measurements or non-additive codes, the authors construct protocols that generate long-range nonstabilizerness while requiring only single-qubit unitary corrections.
In briefUnified algebraic theory of spacetime quantum fault tolerance
SummaryThis paper presents a unified algebraic theory of Clifford spacetime quantum fault tolerance, which supports universal fault-tolerant quantum computation. The authors build a spacetime chain complex from a circuit's tensor network representation, describing fault propagation, detector syndromes, and logical actions. The framework unifies spacetime codes, Gottesman's gadget framework, fault complexes, and ZX calculus, and accommodates phenomenological and circuit-level fault models.
In briefA study of the noise limit for quantum communication through depolarizing channels
SummaryThe paper examines the noise threshold at which a qubit depolarizing channel can still support positive quantum capacity. Analyzing the channel's action on symmetric subspaces of many uses, the authors show that it reduces to a bosonic Gaussian channel. Codes designed for the depolarizing channel can be mapped to this Gaussian channel, optimized more easily, and mapped back, yielding improved lower bounds on the threshold.
Rabsan Galib Ahmed, Sujeet Bhalerao, Sungjai Lee et al.
SummaryThis paper asks when causal loops avoid paradoxes by producing valid probabilities under arbitrary independent local interventions. The authors show that a deterministic classical process is logically consistent if and only if it contains no signaling loops, so cyclic causation is allowed while cyclic signaling is not. They also correct an earlier intervention-free criterion that missed global loops.
Hippolyte Dourdent, Andreas Leitherer, Emanuel-Cristian Boghiu et al.
In briefWhether complex quantum dynamics signals classical hardness in noisy circuits
SummaryThis paper examines noisy Clifford+T and nearest-neighbour matchgate+SWAP circuits to test whether dynamical diagnostics track classical simulability. The authors find the two can separate in both directions: diagnostics may stay nontrivial after classical simulation becomes efficient, or vanish before known efficient algorithms apply. They attribute this to magic and scrambling depending on fourth-order Pauli statistics, while simulation algorithms depend mainly on second-order moments, which local noise suppresses at different rates. The authors conclude that dynamical diagnostics measured on a noisy device do not alone establish classical hardness.
WhyShows noise can decouple complexity diagnostics from classical simulability in both directions, linking this to fourth- versus second-order statistics; novel, but evidence is limited to specific circu
In briefThe Intrinsic Cost of Quantum Syndrome Extraction: a paper on the time and communication needed to extract stabilizer-code syndromes on hardware with a given connectivity graph.
SummaryThe paper defines an intrinsic cut cost χ, computed from the stabilizer group alone and equal to the mutual information across a cut in the maximally mixed code state. It shows that any exact extraction circuit needs at least χ/2 two-qubit operations across the cut, or exactly χ shared Bell pairs when local processing is free. Using this measure, the authors build constant-depth annular surface codes and obtain a tight space–depth tradeoff.
WhyDefines an intrinsic cut cost from the stabilizer group, matches it with lower bounds and constructions, and improves on prior bounds.
In briefQuantum-classical break-even for electronic dynamics of large ring molecules
SummaryThis paper shows that quantum hardware can reach useful accuracy in chemical dynamics simulations within wall-clock times comparable to estimated practical classical computing. The authors target ring-shaped conjugated macrocycles such as benzene and [60]annulene, compressing short-time evolution with tensor networks and replicating the circuit using rotational symmetry. On IBM quantum processors, a hybrid approach was compared against two state-of-the-art classical methods, finding comparable accuracy and computational time under ideal parallelization.
Shu Kanno, Kenji Sugisaki, Takashi Imamichi et al.
In briefScaling laws for quantum networks via a new entanglement transport metric
SummaryThe paper defines entanglement transport, a metric that weights each delivered Bell pair above a fixed fidelity threshold by its endpoint separation, to compare how entanglement distribution scales with node and user counts. It uses a two-layer framework: a resource layer measuring capacity per activation slot and a physical layer measuring service per second. Scaling results are given for honeycomb lattices, homogeneous chains, and spatial expansion.
In briefPolylog-depth quantum thermal state preparation using local recovery maps
SummaryThe paper gives sufficient conditions for preparing quantum thermal states of noncommuting local Hamiltonians in arbitrary fixed spatial dimension with polylogarithmic circuit depth. It combines locality and stability bounds on Gibbs-state reduced density matrices with a high-temperature condition and coarse classical reference Hamiltonians, building the circuit from spatially localized Petz recovery maps. After classical preprocessing, the circuit prepares the canonical purification using N·polylog(N/ε) gates and qubits.
WhyGives polylog-depth thermal state preparation via local Petz recovery, a significant advance, though evidence is mainly theoretical without numerical validation.
Hideaki Hakoshima, Atsushi Iwaki, and Nobuyuki Yoshioka
In briefSeparable decompositions of 2×n states with operator Schmidt rank three
SummaryThis paper studies bipartite states of a qubit coupled to an n-level system whose operator Schmidt rank is three. It shows every such state is a mixture of rank(ρ) pure product states, and in general a mixture of at most n+1 mixed product states. The n+1 bound is tight, as some states need that many terms. The proofs give constructive decomposition methods.
In briefQMA(2) with Limited Shared Entanglement: whether a limited number of shared entangled pairs between two quantum provers changes the power of this quantum proof system
SummaryThe paper studies QMA(2), a quantum proof class in which two computationally unbounded provers submit unentangled quantum witnesses to a polynomial-time quantum verifier. The authors define QMA(2;h), where the provers initially share h EPR pairs. They show that polynomially many shared pairs give no cheating advantage, and that logarithmically many leave QMA(2) unchanged. In a variant that also allows classical communication while preparing witnesses, assuming NP is not contained in BQP, they show that superlogarithmically many shared EPR pairs strictly reduce the model's power.
WhyFormalizes QMA(2) with shared entanglement, proving equality for logarithmic EPR pairs and strict weakening for superlogarithmic ones, a substantive theoretical advance.
In briefPhase-sensitive avalanche sensing of light below shot noise
SummaryThis numerical study examines atomic clouds trapped in photonic cavities. When the cavity mode is tuned to a parity-prohibited transition, the clouds show non-perturbative sensitivity to changes in the cavity field. Under continuous-wave laser pumping, the atomic emissions amplify initially sub-shot-noise fluctuations by about two orders of magnitude, and the amplification is broadly tunable in frequency. The mechanism also preserves the phase imparted by the original ultra-weak light wave.
In briefWeakly measured loops for faster quantum amplitude amplification with unknown success probability
SummaryThe paper proposes quantum amplitude amplification algorithms that interleave Grover rotations with tunable weak measurements in loops, stopping on success and resuming from the post-measurement state on failure. For known target probability, it gives a smaller leading constant in expected Grover iterations than standard Grover and optimised restart Grover. When only a lower bound on the probability is known, it derives the exact expected oracle cost and an optimal measurement strength, with leading constant 1/√2 at the promise boundary. For completely unknown probability, it analyses the critical measurement-strength scale and reports numerically optimised results.
WhyReplaces restarts with weak-measurement loops, giving closed-form expected costs and optimal strengths with improved constants for known and unknown target probabilities.
In briefQuantum Algorithm for Power Grid Layout Optimization Using Branch Exchanges
SummaryThis paper presents a hybrid quantum-classical method for reconfiguring power distribution networks to reduce active line losses. It uses a branch-exchange encoding that guarantees every outcome is a radial network, and a quantum imaginary-time evolution routine guided by a surrogate model trained on AC power flow solutions. Tested on ten cases, the method reached minimal-loss configurations for six and came within 0.2 to 2.18 percent of the reference minimum for the other four.
In briefThermodynamic costs of continuous quantum error correction
SummaryThis paper treats measurement-based continuous quantum error correction as an information engine and analyzes the thermodynamic resources it requires. It separates system-side power from energy transfer and controller-side power from the feedback Hamiltonian. The framework is developed for the one-qubit Ahn-Doherty-Landahl protocol and extended to the three-qubit repetition code. Simulations show that stronger feedback raises steady-state fidelity and lowers conditional-state entropy, but both energy contributions keep growing after fidelity saturates, revealing a tradeoff between logical stabilization and energy cost.
In briefComplexity of Hamiltonians built from one repeated local interaction
SummaryThe paper studies Local Hamiltonian problems where every local term is the same fixed interaction, called the Single-Interaction Hamiltonian (SIH) model. The authors give an exact compiler that turns any finite set of local interactions into a single interaction, yielding a polynomial-time reduction from k-Local Hamiltonian. They also prove QMA-completeness for SIH with 3 and geometrically local SIH with 8, and StoqMA-completeness for stoquastic SIH.
WhyIntroduces an exact compiler reducing k-local Hamiltonians to single-interaction models, with QMA and StoqMA completeness proofs.
In briefExact Maximum-Likelihood Decoding Beyond Treewidth Using Rank-Decomposition Dynamic Programming
SummaryThis paper introduces a decoding algorithm for quantum error correction based on rank-decomposition dynamic programming (Rank DP) that computes exact logical-class likelihoods under stochastic Pauli noise. The authors report arithmetic complexity polynomial in input size and exponential in rank-width, which yields polynomial-time decoding for punctured quantum Reed-Muller codes and a Steane-concatenated family. Numerical experiments show runtime advantages over tested tensor-network implementations on selected instances, and the likelihoods are also applied to learning noise parameters and evaluating decoder optimality gaps.
WhyReplaces tensor-network contraction with rank-decomposition DP, beating treewidth limits, and validates exact likelihood decoding on Reed-Muller codes up to 1023 qubits.
In briefOptimal local recovery cannot change critical orthogonality exponents in quantum spin chains
SummaryThe paper proves a local amplitude-ratio bound for positive ground states and an explicit reverse fidelity comparison under finite-region erasure. An exact lattice calculation for a connected critical Ising chain gives an amplitude exponent of 1/8, which is preserved by any optimal control on a fixed spin set. The authors also present a finite-mediator theorem, a measurement protocol, and counterexamples to stronger inferences.
In briefLayer codes as quantum memories: a 3D error-correcting code tested for syndrome extraction, thresholds and idle robustness
SummaryThis paper evaluates layer codes, three-dimensional local CSS codes built from unrotated surface code patches, as active quantum memories under circuit-level noise. The authors present a decoder-free scheduler for syndrome extraction circuits and compare eighteen layer codes with surface codes of the same distance. A primary family built from a [[4,2,2]] input code reaches a circuit-level threshold of 5.06(7)×10^-3. The layer codes tolerate 1.7 to 3.0 times longer intervals between syndrome extraction rounds than rotated surface codes, at the cost of more operations per unit time and about five times more physical qubits per logical qubit.
In briefEmergent quantum geometric phases in patterned graphene that mimic moiré electronic states
SummaryThe authors fabricate encapsulated monolayer graphene with patterned periodic antidot lattices and measure its magnetotransport. Tuning the lattice size reveals classical commensurability features from cyclotron orbits of similar radii, along with Brown-Zak oscillations from quantum commensurability between the magnetic flux quantum and the lattice unit cell. The paper reports that the system's geometric phase is directly measurable and that lithographic patterning can generate moiré-like spectra without twisting or stacking layers.
Pablo Canteli, Yuriko Baba, Juan Salvador-S\'anchez et al.
In briefMeasuring oscillating signals with one quantum interferometer, using sequential orthogonal mixing
SummaryThe paper proposes a protocol that performs quantum double lock-in detection with only a single quantum interferometry, extracting a signal's amplitude, frequency and initial phase. It uses sequential orthogonal periodic multipulse sequences. With a GHZ input state and two interaction-based operations during interferometry, the authors state that precision for frequency, amplitude and initial phase can approach the Heisenberg limit.
In briefPreparing large cat states efficiently with optical operations
SummaryThe paper proposes an all-optical scheme that uses quantum non-demolition interactions and photon-number measurements to prepare large-amplitude cat states with near-unity probability. No particular photon-number outcome is postselected; every outcome adds to the accumulated photon number, and the protocol repeats until a required threshold is reached. The authors show the scheme is robust to realistic photon loss, the dominant error source in optical systems.
Simon K. Yung, Matthew S. Winnel, Timothy C. Ralph et al.
In briefProof of an energy gap for a Rydberg-blockade model with topological order
SummaryThis paper rigorously proves that a specific Hamiltonian, motivated by the Rydberg platform, has a finite excitation gap. The model contains only two-body blockade interactions and has a topologically ordered ground state in the toric code phase. The authors show the result also extends to a broader class of blockade Hamiltonians realizing non-Abelian quantum double phases. The proof builds on known gap stability results and uses the models' local symmetry.
Simon Fell, Tobias F. Maier, Hans Peter B\"uchler et al.
In briefSpeed limit for randomness in U(1)-symmetric quantum circuits: a two-particle encounter bottleneck
SummaryThis paper studies how randomness forms in quantum circuits that respect a U(1) conservation law, a minimal setting for symmetry-constrained random circuits. It determines the optimal scaling of unitary k-design formation rates across many circuit geometries and finds that for k≥2 the rate is set by a two-particle encounter rate rather than the charge transport rate previously conjectured. The authors prove a matching lower bound and show that symmetry-breaking local gates can circumvent this bottleneck to form designs faster.
WhyShows two-particle encounter rate, not charge transport, governs design formation, disproving a conjecture with a matching rigorous lower bound.
In briefA quantum method for finding the best perfect pairing of graph edges
SummaryThis paper presents a quantum algorithm for the maximum-weight perfect matching problem on general graphs with integer edge weights, running in Õ(n m^{2/3} log W) time, where W bounds the edge weight magnitude. In the dense regime it beats the best known classical combinatorial bound of Õ(m√n log W), making it the first quantum algorithm to asymptotically improve on the best classical combinatorial method for this problem. It builds on the classical framework of Duan, Pettie, and Su, replacing some classical steps with quantum methods and using data structures suited to the QRAM model.
WhyFirst quantum asymptotic speedup over the best classical combinatorial algorithm for weighted perfect matching in general graphs, adapting an existing framework.
In briefFrom wavefunction regularity to eigenvector conditioning: accuracy-cost trade-offs of quantum algorithms for non-Hermitian transcorrelated Hamiltonians
SummaryElectron-electron Coulomb interactions create cusps in electronic wavefunctions, which slow basis-set convergence. The transcorrelated method builds short-range correlation into the Hamiltonian to speed convergence, but the resulting operator is non-Hermitian and non-normal. Using a solvable model with two one-parameter Jastrow correlators, the paper derives energy-error scaling of N^-1/2 for bare projection and N^-3/2 for both transcorrelated versions, where N is the basis size. It argues this faster convergence can give a polynomial improvement in the query-complexity upper bound, and extends the analysis to electronic-structure Hamiltonians.
Cheng-Lin Hong, Emiel Koridon, Paul K. Faehrmann et al.
In briefNoisy shallow quantum circuits in 1D can be classically simulated in parallel, with no quantum advantage
SummaryThe paper studies shallow quantum circuits on a line of qubits with nearest-neighbor gates and independent depolarizing noise before each layer. It presents a randomized parallel algorithm that samples the output distribution to total variation error δ in parallel time 2^{O(d)} log log(n/δ). For constant depth, the authors show the noisy circuit's input-output behavior can be reproduced by a randomized AC^0 circuit, ruling out the unconditional quantum advantage known in two and three dimensions for one dimension.
WhyShows noise splits 1D shallow circuits into independent pieces, yielding parallel classical sampling and an AC0 simulation that rules out quantum advantage.
In briefInitializing silicon spin qubits with microwaves, without reservoirs or measurement
SummaryThis paper demonstrates initialization of a silicon spin-qubit pair in an industrially fabricated Si/SiGe quantum-dot device using a fixed sequence of microwave and baseband pulses, without local reservoir access or measurement-based feedback. Odd spin-parity states relax to the singlet charge state, while blocked even-parity states are microwave-driven and converted to the singlet via singlet-triplet mixing. Repeated cycles give a median singlet-associated charge outcome of 99.4%, and exchange spectroscopy verifies mapping to the target state.
Leon C. Camenzind, Ik Kyeong Jin, Akito Noiri et al.
In briefGenuine Multipartite Nonlocality Is Fermionic Magic
SummaryThe paper shows that free fermion dynamics is classically simulable and that a single non-Gaussian state makes it universal, and that the same boundary governs quantum nonlocality. Measured as fermionic qubits, fermionic Gaussian states reach at most the biseparable bound of the Mermin inequality for three parties and violate the Mermin family by only the factor of one entangled pair for any number of parties, so they are never genuinely multipartite nonlocal. The authors also report experiments on superconducting processors.
WhyProves fermionic Gaussian states are bounded by biseparable limits and links this to computational universality, with superconducting-processor experiments testing predictions.
In briefChiral quantum chaos near exponentially many zero modes in the quantum breakdown model
SummaryThis paper studies a zero-dimensional, all-to-all-interacting version of the quantum breakdown model, a model of randomly interacting fermions motivated by dielectric breakdown physics. It establishes a symmetry classification and quantum chaos features, showing a Z4 periodicity in the number of fermionic modes and realization of all five chiral symmetry classes. The model hosts exponentially many many-body zero modes protected by the chiral index, and the authors analyze the scaling of the spectral gap and symmetry-enriched hard-edge spectral statistics as signatures of chaos.
In briefSquare-root growth of operator entanglement in an integrable circuit
SummaryThe paper builds a four-state brickwork circuit whose gate is a permutation matrix solving the constant Yang-Baxter equation, as a counterexample to the expected logarithmic growth of operator entanglement. The von Neumann operator entropy grows as a square root of time, while some Rényi entropies grow linearly or logarithmically depending on their index. Matrix product operator simulations at fixed error below one need bond dimension growing at least exponentially in the square root of time.
WhyConstructs a counterexample to logarithmic operator entanglement growth, backed by an exact asymptotic formula and matrix product operator lower bounds.
In briefLogical operator decomposition for distance analysis of bivariate bicycle codes
SummaryBivariate bicycle (BB) quantum codes are a prominent family of finite-length quantum LDPC codes whose minimum distance is usually found numerically. The paper decomposes the Z-logical quotient into a short exact sequence, yielding an explicit logical basis, a dimension formula, and a componentwise distance identity. It proves that both components have rank k/2 for every BB code. Using a translation-anchored cluster search, the authors establish distances for six standard BB codes of lengths 18 to 288 and enumerate all minimum-weight logical operators.
WhyDecomposes BB code logical operators via a short exact sequence, proves a componentwise distance identity, and exactly determines six standard code distances with exhaustive enumeration; novel and ver
In briefA quantum battery charged selectively by frequency through an anharmonic mediator
SummaryThe paper studies a finite-time charging scheme in which a driven two-level system transfers nonequilibrium resources to an interacting two-qubit battery via a weakly anharmonic three-level mediator. The mediator's lower transition enables nearly complete single-excitation transfer, while a two-excitation cascade charges the doubly excited battery state with a modest efficiency reduction. Detuning scans quantify spectral tolerance, and the full charge-discharge cycle's energy balance closes to numerical precision.
Sujay Mondal, Anumita Mukhopadhyay, Siddhartha Dutta et al.
In briefEfficient learning of fermionic magic states under free-fermion evolution
SummaryThis paper shows that a family of fermionic magic states can be learned efficiently under unknown number-conserving free-fermion evolution. The method uses the spectral structure of particle reduced density matrices to separate individual block contributions and reconstruct the hidden block structure. For a fixed per-block particle-number bound, it uses single-copy measurements and polynomial resources to recover a compact state description with prescribed fidelity.
In briefInjecting states into quantum codes with constant overhead
SummaryThis paper presents the first known fault-tolerant scheme for injecting bare physical qubits into quantum error-correcting codes, and ejecting them back, with constant space and time overhead. The procedures succeed under circuit-level locally stochastic noise, corrupting each qubit only with a small constant probability. All gadgets run in constant circuit depth, and the qubit count grows linearly with the number of logical qubits.
WhyGives the first constant-overhead fault-tolerant injection and ejection scheme, resolving an open problem with detailed proofs and broad impact on quantum error correction.
In briefComplete set of product states for the three-qubit Shifts witness and its optimality
SummaryThis paper studies the three-qubit Shifts unextendible product basis and determines all complex product states at which its projector witness reaches equality. It shows there are exactly eight product rays, which span the full three-qubit Hilbert space, unlike four previously known ones. This establishes that the witness satisfies the spanning criterion and is optimal. As a corollary, for a white-noise family it detects entanglement when the noise parameter is below about 16.29%.
In briefInformation Potential: A New Method for Bounding Quantum Communication
SummaryThis paper introduces a variational technique called the information potential for analyzing quantum information complexity (QIC) and proving quantum communication lower bounds. The method is inspired by the resolvent representation of quantum relative entropy. The authors obtain optimal lower bounds for the QIC of the two-bit AND function and for r-round Set-Disjointness, answering an open problem. They also prove a direct-sum theorem and a nearly tight tradeoff for asymmetric Set-Disjointness.
WhyIntroduces a new variational information-potential method yielding optimal lower bounds for AND's QIC and r-round Set-Disjointness, plus a direct-sum theorem and tradeoff, with strong rigorous support
In briefA dichotomy result for MIP* under unital noise
SummaryThis paper studies two-prover one-round quantum proof systems in which the provers share EPR-type states affected by unital noise. The authors show that if the state is separable or has quantum maximal correlation below 1, the complexity class equals NEXP (MIP); if it is entangled with maximal correlation equal to 1, the class equals RE. The paper also introduces positivity-preserving low-degree approximation and a dimension-reduction technique called Pauli folding.
WhyGives a complete dichotomy for unital-noise MIP* with new positivity-preserving low-degree approximation and Pauli folding techniques, with rigorous proofs.
In briefSQD-Agent: an LLM agent that turns plain-language quantum chemistry goals into runnable hybrid quantum-classical workflows.
SummaryThis paper introduces SQD-Agent, a framework that uses large language models to convert natural-language user requests into executable workflows for quantum chemistry. The workflows use algorithms from the Sample-Based Quantum Diagonalization (SQD) family. The modular design supports multiple quantum backends and classical solvers, and includes profiling, caching, and convergence visualization. The paper also covers error mitigation on real quantum hardware and an analysis of candidate mitigation schemes in terms of error recovery and computational budget.
In briefTunable-Resource Quantum Circuits for Nonlinear Element-Wise Transformations
SummaryThis paper presents a resource-tunable quantum framework for implementing nonlinear element-wise transformations. It decomposes a (2^d-1)-degree polynomial into m lower-degree factors, where m controls the degree of parallelism. This trades off query complexity, extra gate complexity, ancilla count, and normalization, with a construction of depth O(n+d) matching the query lower bound. The authors also analyze the effects of polynomial approximation and implementation errors, and demonstrate the framework on sigmoid and tanh activations and pixel-wise image transformations.
In briefLimits on adiabatic path geometry from gap-width classes
SummaryThis paper studies how the geometry of the adiabatic ground-state path, including its length and total curvature, depends on the energy gap profile of a Hamiltonian. The authors define a width class from how the region where the gap is small shrinks, which yields tighter bounds on path length and curvature and on evolution time. The bounds are demonstrated on adiabatic Grover search, the XXZ spin chain, and molecular electronic Hamiltonians.
In briefOptimal learning of quantum states and channels that respect symmetries
SummaryThe paper presents collective tomography protocols for quantum states and channels with known symmetries, reducing learning to pure-state estimation via random purification and dilation. It gives optimal copy complexity for covariant states, query scalings for covariant channels, and optimal query scalings for permutation-covariant channels under diamond and Choi trace distances. The authors also construct efficient quantum circuits approximating the Hayashi measurement, resolving an open problem in quantum tomography.
WhyGives optimal learning bounds for covariant states and channels and resolves an open problem on efficient Hayashi measurement circuits, with strong theoretical support.
Satoshi Yoshida, Kazuki Okigami, Pietro M. Posta et al.
In briefMultipartite entanglement hidden in spin-chain vector-chiral correlations
SummaryThis paper derives a rigorous entanglement bound for the total nearest-neighbour vector chirality of periodic spin-1/2 chains with an even number of sites. Exceeding the bound on quantum Fisher information per spin certifies an entanglement depth of at least k+1. Using infinite DMRG and thermal pure quantum-state calculations, the authors show multipartite entanglement in vector-chiral correlations at zero and finite temperatures in an extended J1-J2 XXZ chain.
In briefBenchmarking noisy random quantum circuits by measuring their response to controlled perturbations, without computing ideal output probabilities
SummaryThe paper proposes a metrological benchmark for random quantum circuits based on quantum Fisher information (QFI), which measures a circuit's response to controlled perturbations. It relates QFI to out-of-time-order correlators and compares Haar-random and Clifford circuits, finding that Clifford circuits yield zero QFI despite extensive operator spreading. For noisy implementations, the authors derive an exact relation between noisy and ideal QFI under a global white-noise model..
In briefTwisted convolution identity and ghost r-SICs from finite quantum dilogarithms
SummaryThis paper extends Radchenko and Wheeler's proof of a finite pentagon relation, originally shown for rank-1 cases, to all rank-r admissible tuples. It also gives a dictionary relating the modular quantum dilogarithm to the Shintani-Faddeev modular cocycle conventions used in the two papers. The authors conclude that when r < (d-1)/2 and (d²-1)/(r(d-r)) is an integer, there exist ghost r-SICs, which are configurations of d² rank-r subspaces in C^d satisfying a non-Hermitian equichordal condition. Under the Stark conjecture, these are Galois conjugate to Hermitian r-SICs.
In briefEfficient electron overlap integrals for Gausslet orbitals on cubic lattices
SummaryThis paper presents efficient algorithms for computing kinetic, nuclear, and electron repulsion overlap integrals for Gausslet orbitals, a localized and systematically refinable basis set on a cubic lattice. The methods exploit translation, permutation, and octahedral symmetries, reorder nested sums, and truncate small coefficients early. On a 5×5×5 grid, symmetries reduce the electron repulsion integrals from 244140625 to 138187, and the remaining integrals run in under one second on a laptop at a 10^-5 truncation tolerance. The authors demonstrate the approach by computing the ground states of the hydrogen atom and molecule.
Xianrui Yin, Fereshteh Ghasempour, Christian B. Mendl
In briefComparing Hybrid Quantum and Classical Self-Supervised Learning for Fingerprint Recognition at Equal Width
SummaryThis paper tests whether adding the QuFeX quantum feature-extraction module to self-supervised learning improves fingerprint recognition. Three frameworks (SimCLR, MoCo v2, BYOL) are compared with their classical versions at matched width of 8 features, using the SOCOFing dataset and CIFAR-10 as a control. Single runs favor the hybrid for both contrastive methods, but a multi-seed analysis of BYOL shows no reliable difference, suggesting the benefit depends on the learning objective.|
Maria S. Edwards, Kidwell Dlamini, Pin-An Lin et al.
In briefQuantum frozen-Oseen homotopy method with LCHS for nonlinear PDEs
SummaryThis paper proposes a Frozen-Oseen quantum homotopy method (FOQHAM) that turns nonlinear PDEs into a fixed affine linear system suitable for quantum linear evolution. The system is propagated using Linear Combinations of Hamiltonian Simulations (LCHS), without outer homotopy iterations or profile updates. Classical numerical tests on Burgers, KdV, Zakharov-Kuznetsov, and one- and two-dimensional isothermal compressible Navier-Stokes equations show accurate non-iterative approximations in the tested regimes.
In briefGeometric analysis of bound entanglement in bipartite qutrit systems
SummaryThis paper studies bound entanglement in bipartite qutrit (three-level) quantum systems by combining the PPT criterion with a non-completely positive map. It analyzes a three-parameter family of states and gives an exact analytical and geometric partition of the physical state space. The authors derive the quadratic boundary of the PPT region and a linear threshold identifying the PPT bound entangled sub-region, which they state accounts for 14.76% of the PPT leaf area.
In briefAn uncertainty principle for entanglement in quantum states
SummaryThis paper studies two orthonormal bases of a bipartite Hilbert space, one with weakly entangled states and one with strongly entangled states. The authors show that a state localized in one basis must be delocalized in the other, and give a lower bound on the sum of Shannon entropies of its coefficients in the two bases. The result also limits how weakly entangled states can represent ground states of Hamiltonians whose energy eigenstates are strongly entangled, and it is illustrated numerically for the Sachdev-Ye-Kitaev model.
In briefMetachecks in BB codes: how well redundant checks catch measurement errors
SummaryThis paper studies bivariate bicycle (BB) quantum error-correcting codes, whose dependent stabilizer checks create extra parity constraints called metachecks. It asks how far these constraints can identify faulty syndrome measurements and when some ambiguity remains unavoidable. Exact calculations on finite codes, including a 72-qubit BB code and the 144-qubit Gross code, illustrate the regimes. Simulations indicate that joint data-measurement decoding is more robust than a separated repair stage on the more ambiguous codes.
In briefComplete parameterization and topology of parameter space for discrete Wigner representations on d×d phase space
SummaryThis paper completely parameterizes and topologically classifies single-qudit discrete Wigner representations on a d×d phase space that are Hermitian, unit-trace, Hilbert–Schmidt orthogonal, and Weyl–Heisenberg covariant. It determines the independent parameter freedom in every dimension. The parameter spaces differ between odd and even d, with explicit circle (S¹) and discrete (Z_d) factors given for each case.
In briefOptimal Testing of Fermionic and Bosonic Gaussian Quantum States
SummaryThis paper studies property testing of quantum states, deciding whether a pure state belongs to a class or is far from it. For five classes of Gaussian states, including fermionic and bosonic Gaussian states and coherent states, the authors prove a tight sample complexity of Θ(1/ε²) that does not depend on system size. The tester projects two or three copies onto the top irrep, and matching lower bounds come from binary state discrimination. The results also extend to the tolerant setting.
WhyProves size-independent Θ(1/ε²) sample complexity for five Gaussian state classes with matching lower bounds, a notable theoretical advance with a narrow scope.
In briefQuantum circuit compilation without extra overhead
SummaryThis paper studies the overhead incurred when compiling quantum circuits from a continuous gate set into a discrete universal gate set {H, T, CNOT}. Typically each original gate is replaced by O(log(G/ε)) elementary gates. The authors show that, using adaptivity, this overhead can be avoided for polynomial-sized circuits with inverse-polynomial target error, and state that the result is optimal.
WhyUses adaptivity to remove the logarithmic per-gate compilation overhead, proving optimality, a substantial theoretical advance with mainly theoretical scope.
In briefEngineered dissipation for steering thermal states through a quantum critical point
SummaryThe paper proposes a Markovian dissipative framework that uses engineered population currents with non-negative transition rates to steer a quantum system along a prescribed thermal trajectory in finite time. It is applied to the transverse-field Ising model driven across its quantum critical point. Compared with unitary dynamics and a natural thermal bath, the engineered approach strongly suppresses thermal excitation deviations and offers control over coherence damping via a free common current. The residual deviation scales more steeply with driving speed than the Kibble-Zurek behavior seen under unitary evolution.
In briefBifurcation limits on signal-to-noise gains in nonlinear exceptional-point sensors
SummaryThe paper examines whether nonlinear exceptional-point sensors can achieve divergent signal-to-noise gains. The authors prove that any singular stationary frequency response requires a fixed-point bifurcation of the nonlinear dynamics. In local fluctuation regimes, responsivity and frequency uncertainty singularities cancel exactly, and a path-information bound limits the long-time SNR enhancement.
In briefProvable classical and quantum local algorithms for Max-k-Cut, showing quantum advantage at moderate girth
SummaryThis paper studies local classical and quantum algorithms for the Max-k-Cut problem on d-regular graphs with girth g. The authors propose a local vector algorithm based on the Thompson, Parekh, and Marwaha vector construction, which gives the best provable cut-fraction guarantee among known efficient classical algorithms on regular graphs for k≥3. They also develop tensor network methods to analyze QAOA performance and compute it in the infinite-degree limit. For depths p≥9, corresponding to girth g≥20, QAOA achieves a better guaranteed cut fraction than the classical algorithm, which the authors present as an apparent quantum advantage.
In briefEntropy threshold: a simple estimate of quantum error-correction performance
SummaryThe paper proposes a simple entropy-based proxy that estimates thresholds for various quantum error-correction protocols, from code-capacity settings to circuit-level noise models. It locally compares noise entropy with the error information gained from stabilizer measurements. Although it neglects stabilizer-outcome correlations and code degeneracy, the proxy captures main trends and yields threshold estimates in good agreement with numerical simulations.
In briefConstant-rate codes with fully addressable transversal T gates and sparse checks
SummaryThis paper extends Wills' work by constructing asymptotically good CSS codes with fully addressable transversal T gates, allowing selective control over which logical qubits receive a T gate. The authors also present Pauli-LDPC codes with transversal T whose distance grows as the rate vanishes, and a protected-subsystem construction with constant protected rate, distance Ω(√n), and sparse non-Pauli checks.
WhyExtends Wills' transversal-T work with asymptotically good CSS codes and fully addressable constructions, offering new theoretical results with distance bounds, though practical impact is limited.
In briefQuantum pseudorandom codes: error-correcting codes that look random to quantum observers
SummaryThe paper introduces quantum pseudorandom error-correcting codes (QPRCs), whose encodings are computationally indistinguishable from Haar-random isometries or the completely depolarizing channel. Under the assumption that Learning Parity with Noise is hard for quantum algorithms, the constructions are robust to local quantum noise. The authors also present a new efficient decoding procedure within the codeword-stabilized framework for nonlinear classical codes.
WhyIntroduces quantum pseudorandom error-correcting codes under an LPN assumption and resolves an open decoding problem for nonlinear CWS codes, with substantial theoretical support.
In briefEigenstate thermalization near the ground state when many local symmetries are mutually incompatible
SummaryThis paper studies how an extensive number of mutually incompatible local symmetries affects eigenstate thermalization (ETH). Using bond-dependent Z2-symmetric spin-1/2 Hamiltonians and spectral diagnostics, the authors show that non-integrability can survive despite many conserved quantities when they do not commute. They further show that ETH behavior extends deep into the many-body spectral edge, near the low-excitation-energy region close to the ground state.[]
In briefCommuting Local Hamiltonian Problem: Relativized Evidence Against BQP-Hardness
SummaryThis paper studies the commuting local-Hamiltonian (CLH) problem, a restriction of the local-Hamiltonian problem in which all terms must pairwise commute. The authors introduce a classical-oracle version of the class QIMA and construct a classical oracle under which BQP is not contained in QIMA, so QIMA and QMA differ relative to that oracle. This provides relativized evidence that the general CLH problem is not BQP-hard, and therefore not QMA-complete.
WhyIntroduces a classical-oracle QIMA and builds an oracle separating BQP from QIMA, giving relativized evidence about CLH hardness.
In briefFrequency comb generation in active optomechanical cavities: numerical study of a VCSEL coupled to mechanical modes of a suspended mirror
SummaryThe authors numerically study an electrically pumped vertical-cavity surface-emitting laser coupled to two mechanical modes of a suspended mirror. Displacement-dependent cavity loss and radiation-pressure feedback link mechanical motion to carrier and photon densities, producing harmonics and combination-frequency components in the photon-density spectrum. Mechanical quality factors mainly set comb line intensities, while resonance frequencies determine their spectral distribution. Initial mechanical amplitudes affect threshold current and relative spectral intensities, and exciting both modes together yields a broader set of components.
In briefQuantum-echo Markov process for combinatorial optimization
SummaryThis paper introduces a quantum-echo Markov process for combinatorial optimization, using quantum annealing or QAOA dynamics to build the transition kernel. Longer annealing times or more QAOA layers let transitions reach distant configurations while limiting large energy changes. The authors propose quantum-echo local optimization, with and without greedy descent, and find that adding greedy descent substantially improves performance.
In briefDefect-aware parallel atom reloading protocol for neutral-atom quantum computers
SummaryThis paper proposes a protocol that reloads atoms into defect positions, where atoms have been lost, based on defect information to raise the atom filling rate. It combines a coherent reloading operation that preserves quantum information with a planner that chooses reloading patterns. In low-noise numerical experiments on a 36 x 90 atom array, the average filling rate rose from 98.61% to 99.94%, and each reload was planned in under 0.1 ms.
In briefBoundary-rank limits and measurement-assisted recovery in dissipative flat-band preparation
SummaryThe paper studies why number-conserving dissipative cooling can fail to prepare an interacting flat-band target when Pauli exclusion blocks transfer destinations. It analyzes this failure on vertex-edge decorated graphs, relating residual flat source directions to the rank of a Gram-matrix cut block, and constructs a finite-range chain example with fidelity bounds. The authors show that dephasing the occupations of compact bright modes can restore global attraction to the ferromagnetic target under stated conditions, and support the analytic results with Gram calculations, finite-system dynamics, and Liouvillian spectra.
In briefOptical study of negative g-factor dopants in an antiferromagnet, Nd:GdVO4
SummaryNeodymium-doped gadolinium vanadate (Nd:GdVO4) is a common room-temperature laser material. This paper studies its optical properties below the 2.5 K Néel temperature and observes narrow optical transitions. The authors show that rare-earth ions in magnetic hosts can reveal the sign of magnetic g-factors, and report clear evidence of negative g-factors from the interplay of exchange and Zeeman interactions.
Lachlan W. Cronin, Masaya Hiraishi, Luke S. Trainor et al.
In briefClassical Algorithms for Computing Functions in Gaussian Boson Sampling
SummaryGaussian boson sampling (GBS) uses squeezed inputs and passive linear optics to generate photon-number patterns, often studied for quantum advantage. This paper examines tasks that apply functions to these photon-number outcomes. By analyzing the irreducible decomposition of fixed-photon-number operator spaces, the authors prove that such function expectation values, averaged over passive linear-optical networks, can be classically evaluated for finite squeezing strength. They also give a classical algorithm that estimates this value to inverse-polynomial additive error.
WhyAnalyzing irreducible decomposition of fixed-photon-number operator spaces, it proves average-case function expectations are classically computable with an inverse-polynomial error estimator, substant
In briefQuantum quenches of AKLT scar states via Clifford-augmented tensor network simulation
SummaryThis paper uses a Clifford-augmented simulation method to study the ground and scar states of the AKLT model, finding they have high non-stabilizerness (magic) as measured by stabiliser Rényi entropy. Their low entanglement still allows efficient representation as low bond-dimension Clifford-augmented matrix product states, which are time-evolved with the Clifford-augmented TDVP algorithm. After a quench within the Haldane phase to the antiferromagnetic Heisenberg Hamiltonian, the string order in the scar states decays, more strongly with increasing bimagnon number.
In briefJoint measurability of coplanar quantum measurements
SummaryThe paper studies unbiased qubit positive operator-valued measures (POVMs), which can be uniquely described by Bloch vectors. It proves that POVMs with coplanar Bloch vectors are jointly measurable if and only if the perimeter of the convex hull of those vectors is at most 4. As a corollary, it affirmatively answers a conjecture by Andrejic and Kunjwal (2020) that POVMs with Bloch vectors on the same circle admit a joint measurement device.
In briefLower bounds on convergence speed for optimistic matrix mirror-prox in quantum zero-sum games
SummaryThis paper examines whether the accuracy dependence of optimistic matrix mirror-prox (OMMP) for quantum zero-sum games is tight, and whether last-iterate convergence can be guaranteed. Using explicit one-qubit-per-player games, it proves an Ω(1/ε) lower bound for the uniform-average output. It also shows that optimistic gradient descent-ascent and optimistic matrix multiplicative weights updates converge at rates Θ(1/t) in distance and polynomially in duality gap, respectively.
Yiheng Su, Emmanouil-Vasileios Vlatakis-Gkaragkounis, Pucheng Xiong
In briefRelative entropy of entanglement and tripartite minimal surface: a holographic dual for three-way boundary entanglement
SummaryThe paper proposes a holographic dual for the relative entropy of entanglement of a tripartition A:B:C of the boundary, given at leading order by the area difference between a minimal bulk tripartition surface and the minimal surface homologous to AB, divided by 4G_N. For general random tensor networks an upper bound is proved, and a matching lower bound is established rigorously for networks of one, two, and three Haar random tensors. The authors note that in suitable geometries the minimal tripartition surface can form a tri-junction resembling the Mercedes logo.
In briefEntangling atomic quantum memories with high-order modulated coherent states
SummaryThis paper proposes distributing multi-ebit entanglement between cavity-coupled atomic memories using high-order coherent-state modulation and collective midpoint measurements. An SRM-inspired 16-QAM receiver reaches 1.75 ebits per network mode use at 0.5 dB end-to-end loss, beating single-qubit-per-mode benchmarks by 5.8 dB. The advantage holds for per-interface loss below 0.2 dB and phase errors below 0.2π.
In briefConditions for preserving observable records when hidden states update in retrocausal models
SummaryThe paper studies whether hidden-variable models keep observable records consistent when devices are connected or reused. It separates two composition requirements, characterized by an invariant space for continuation mass and an observation space for indistinguishability of hidden preparations. Finite-state and wrapped-Cauchy spin examples test these criteria, which apply only to specified instrument families and accessible records.
In briefBreaking the Shot-Noise Barrier: Cached Recycled Variance-Reduced Gradients for Quantum Optimization
SummaryVariational quantum algorithms depend on classical optimizers, but parameter-shift gradients need circuit runs that scale with the number of parameters, which is costly for deep circuits. Simultaneous perturbation needs only constant runs but has high gradient variance. The paper introduces CRVG, a variance-reduced gradient method in non-recursive and recursive variants, which caches center-point measurements to cut inner-loop circuit cost by 25%. Tests on MaxCut and Maximum Independent Set benchmarks show CRVG helps in specific settings, while standard simultaneous perturbation remains a robust baseline across depths.
In briefClique optimization via an entropy-computing photonic platform
SummaryThe paper proposes a framework that uses the Motzkin-Straus theorem to turn discrete clique problems into sum-constrained continuous quadratic programs, solved on QCi's Dirac-3S photonic entropy computer. Extensive benchmarks on the DIMACS suite are reported. The platform matches or beats two independent classical baselines on more than four-fifths of instances. Well-tuned classical continuous optimizers keep an advantage only on the hardest planted-clique instances.
In briefA paper on experimentally testable quantum advantage in shallow circuits
SummaryThe paper refines the finite-size classical soundness bound of Aasnaess's graph-distributed construction so that it depends linearly on the number of players. Combined with disjoint-player repetition, it yields a two-round test on a single processor for any finite nonlocal game with a finite-dimensional perfect quantum strategy and classical winning probability below 1. The quantum implementation has constant depth per round and always wins, while a classical device's winning probability is bounded. An explicit proposal for an experimentally testable quantum advantage using 99 qubits is also presented.
In briefRayleigh-Wood threshold controls photon correlations in atomic arrays
SummaryThis paper studies two-dimensional atomic arrays, single atomic layers that couple photons to collective lattice modes in free space. The authors show that when the lattice spacing crosses the diffraction threshold (the Rayleigh-Wood anomaly), newly opened radiative channels broaden the dominant collective modes and turn specular reflection into diffuse off-axis scattering. They also report that in transmission, the connected two-photon contribution drops by several orders of magnitude once the Rayleigh channels open, while staying near unity in reflection.
In briefProof that the generalized semi-Clifford conjecture holds for quantum gates up to level 4
SummaryThe paper studies the Clifford hierarchy, which classifies quantum gates that can be implemented fault-tolerantly via gate teleportation. It addresses the generalized semi-Clifford conjecture, which states that each gate is, up to Cliffords, a permutation times a diagonal matrix. By extending fixed-point arguments to conjugation groups, the authors prove the conjecture for levels k ≤ 4 in any prime dimension d.
Maxwell Marcus, Sathyawageeswar Subramanian, and Marcel Dall'Agnol
In briefGeometric estimation of NV charge-state contributions via low-dimensional spectral representation
SummaryThis paper presents a geometric method for estimating the photoluminescence contributions of neutral (NV0) and negatively charged (NV-) nitrogen-vacancy centers in diamond. Spectra acquired at different excitation intensities are mapped into a low-dimensional space, and zero-phonon-line information from two spectra calibrates the trajectory. The NV- contribution is then determined by geometric projection and compared against an independent dual-excitation reference method.
In briefGenerating Random Unitaries with Conjugated Hamiltonian Evolutions
SummaryThe paper proposes generating random unitary operations through a sequence of evolutions under a fixed Hamiltonian conjugated by a global control operator. Using a geometric argument, the authors show that under uniform parameter sampling, the generated unitaries converge exponentially to the uniform Haar measure on the accessible unitary group beyond a bounded sequence length. This holds when the connected Lie subgroup generated by the conjugated Hamiltonians is closed. The approach is presented as applicable to platforms such as Rydberg atoms, optical-lattice bosons and fermions, transmons, and trapped ions.
Oscar Scholin, Apollonas S. Matsoukas-Roubeas, Sathyawageeswar Subramanian
In briefQuantum query complexity of searching a linked list
SummaryThis paper studies how many quantum queries are needed to find a marked element in a linked list, in a query model where the address space size N matters. Both the decision and search versions have quantum query complexity Θ(min{ℓ, (Nℓ)^{1/4}}), where ℓ is the list length. When N is smaller than ℓ³, this is strictly less than the Θ(ℓ) cost of ordinary traversal. The same tight bounds also hold for doubly linked lists.
In briefExact diagonal completion for QAOA placement circuits
SummaryThis paper studies how unused encoding states can simplify quantum circuits for QAOA applied to placement, by exactly completing unspecified diagonal-operator entries without changing ideal results. The authors build exact Manhattan-distance operators via weighted-ℓ1 optimization of Walsh coefficients. Across 160 geometries, this completion reduced CX counts in all 96 cases with unused binary codes under Gray-code synthesis, though reductions depend on the compiler. The authors state that no end-to-end advantage is established.
In briefExtractable work from energetic coherence and the third law
SummaryThe paper studies a qubit battery charged by repeated collisions with clusters of thermal ancillas, generating steady-state coherence in the energy basis. It gives closed-form expressions for coherence and ergotropy, showing that for large clusters ergotropy becomes independent of bath statistics. At low temperatures the ergotropy plateaus, which the authors link to the third law of thermodynamics.
O. S. Y\'a\~nez-Soria, G. Guarnieri, B. \c{C}akmak et al.
In briefWhen Error Mitigation Backfires: Budget-Aware Evaluation and Calibration Limits
SummaryThis paper shows that quantum error mitigation can sometimes make results worse rather than better. In simulations with control-offset miscalibration, unconstrained zero-noise extrapolation estimators performed worse than no mitigation on 38-63% of instances and produced extreme tail errors. The authors also compare methods at equal online shot budgets and examine whether provider-reported calibration data reflects the device state at execution time.
Guilin Zhang (Workday AI Research), Kai Zhao (Workday AI Research), Xiquan Cui (Workday AI Research) et al.
