A theoretical note separates what Pontryagin duality and Fourier analysis provide from the physical choice that identifies a mathematical coordinate with momentum.
A mathematical study reports finite-dimensional quantum channels whose two-copy minimum output entropy eventually falls below twice the one-copy value, while leaving the related classical-capacity question open.
A theoretical preprint presents a way to turn impedance and admittance responses into exact Hamiltonians and controlled reduced models for superconducting circuits. Its examples cover discrete filters, finite-band metamaterials, nonreciprocal waveguides and giant atoms, while stressing that the reductions depend on their approximation regimes.
A preprint study found that disagreement between a fast decoder and its syndrome-consistent correction can identify shots likely to fail, focusing deeper computation where it helped most in simulation.
A preprint describes recursive quantum-code constructions that increase code distance and certify circuit-level protection, while simulations show steeper suppression under a specified setup.
A preprint presents PDBQITE as a shallower alternative to recursive DB-QITE, with reported gains in molecular and graph simulations but an exponentially declining chance of completing many postselected steps.
A superconducting quantum thermal machine recorded a TUR ratio below the classical benchmark at 81 mK, while the result remains a preprint and relies on inferred photon fluctuations and model comparisons.
An exact theoretical study of a symmetric double-Morse oscillator distinguishes changes in the modeled ground state from changes caused by how the same quantum state is represented.
The proposed method transfers polynomial quantum states between source and target systems by relative decoding, but its benchmark result remains a limited comparison.
A construction-aware scheduling method uses edge symmetries built into structured qLDPC codes to shrink the circuit-design problem. Tests reported lower-bound or near-lower-bound CNOT circuits across Lifted Product, Balanced Product and Quantum Tanner examples.
A reinforcement-learning protocol adapted each simulation step and reported lower conservation-law errors than fixed-step and adaptive baselines in classical tests.
A theoretical study maps how a single photon carries magnetic-field information after scattering and finds that frequency-resolved counting can approach the model's full quantum-information benchmark under selected conditions.
A hybrid quantum-classical system was judged stronger than three classical generators on synthetic MDS data in a seven-variable proof of concept. The reported result concerns model-quality metrics, not clinical outcomes.
Researchers used fiber-distributed, time-bin-entangled photon pairs to run a Bell test with random measurement settings. The reported result addressed locality, setting choice and postselection, but not the detection loophole.
A preprint models a multimode cavity, a Bose-Einstein condensate and a trapped two-level atom as a tunable Unruh-DeWitt detector. Its calculations report stronger response at low uniform linear acceleration, while circular motion shows a different, nonthermal pattern.
A theoretical quantum-information study finds that exact localization of one changing interval has the same large-sequence success limit under the optimal collective measurement and a square-root measurement, across several interval-length settings.
A mathematical analysis links equal-weight magic-witness thresholds in codeword-stabilized quantum codes to the affine geometry of their classical word sets. It gives exact lower bounds for selected examples and a family-level formula, but leaves the larger Rains family without a scalable algorithm.
A study compares small entangled photon states in a lossy multipass interferometer and reports that rotated NOON outperformed NOON and BAT, while marginally outperforming flat in the stated settings.
An ideal state-vector benchmark found that the best quantum encoding changed with the dataset, and that dense circuits sometimes fit training data far better than test data.
A computational quantum study combines operator backpropagation with two measurement protocols and finds that the lower-shot choice depends on the circuit and observable set.
A methods preprint compares decomposed and naive quantum-walk circuits, with lower model-based CNOT upper bounds in selected one-dimensional and torus-graph examples.
A mathematical study of quantum encodings finds an exact worst-case margin in a dense Majorana model, while broader bounds show how decoding and measurement costs enter the tradeoff.
A laboratory study reports that two molecules embedded in helium nanodroplets kept rotating beyond expected centrifugal limits, while their alignment signals pointed to markedly different effective thermalization times.
A theoretical study identifies a direct-basis and equatorial-basis analyzer pair that jointly screens two multipair false-herald classes, then compares three- and four-source networks with matched all-BSM designs under modeled loss and multiplexing.
A laboratory experiment used entangled photons and a polarization-coupling metasurface to reveal several measured topological states, including up to three states selected from one device.
A theoretical model treats temperature as part of the mathematical description of spin transport in a ferromagnet. It finds only a slight change in the proposed gauge potential across the tested temperatures, while the modeled magnetization pattern changes more visibly.
A theoretical preprint associates competing coordinate-like and momentum-like soft modes with tunable multicritical structures, with distinct signatures in excitation stability and ground-state responses.
A computational study found that activation-selection networks could fit parameter-dependent quantum energy levels with compact formulas, while fixed-basis quadratic regression remained as accurate or slightly better.
A laboratory demonstration on one transmon from a 64-qubit processor reported a 97(1)-nanosecond operational measurement duration, 0.17(1)% two-state assignment error, and mean leakage of 2.7(2) × 10^-5 per measurement at the assignment optimum.
A preprint describes a five-qubit fluxonium-resonator-fluxonium processor with connectivity-four wiring. It reports near-coherence-limited single-qubit gates, a refocused two-qubit result with no measured spectator-state dependence beyond statistical uncertainty, and a five-qubit GHZ fidelity of 90.0(2)%.
A computational study of disordered Ising spin networks found that forecasting precision and collective-relaxation memory scaled fastest near the onset of information scrambling, while the effect depended on forecast horizon and remains untested on hardware.
A theoretical arXiv analysis identifies a conditional peak in information gained per photon absorbed by an object. Within its fixed schemes, one recycled photon is favored, but adaptive strategies can lie beyond the reported ceiling.
A new preprint reports that an approximate quantum-error decoder closely matched exact maximum-likelihood decisions in several simulations, but larger-code and hardware tests revealed clear boundaries.
A field test paired atom-based gravity sensing with inertial navigation without GNSS and recorded repeatable maritime gravity profiles, including anomalies at roughly 300-metre along-track scale.
A laboratory characterization found 0.06 input-referred photons of added noise in phase-sensitive operation, compared with 0.71 photons in phase-preserving operation.
A one-dimensional open-system simulation found finite clusters that retained quantum coherence. Adding nearest-neighbor repulsion was associated with numerical evidence compatible with genuine phase separation, but the extended model was too small for reliable fluctuation scaling.
A new arXiv preprint proposes using a mixed-state geometric phase to track how two modeled quantum oscillators reorganize as their dynamics approach synchronization and critical regimes. The analysis finds distinct response patterns in the models, but the proposed measurement remains untested.
A mathematical preprint describes hybrid quantum networks whose modeled output probabilities use matrix functions beyond determinants and permanents, with quasiparticle expressions that include non-permutation terms.
A computational study found that adding a modeled fault-tolerant √T injection primitive was associated with lower average costs in a finite, noisy single-qubit synthesis benchmark, despite the primitive’s higher assigned cost.
A mathematical study of embedded observers in a deterministic classical model finds that limits on learning can arise even without quantum mechanics. The result depends on what an observer can record, what information is available before an interaction, and whether the goal is to reconstruct the past or predict the future.