In briefQuantum error correction in near-critical toric codes
SummaryThis paper studies error correction for topological codes pushed toward a quantum phase transition, focusing on quantum information stored in the transverse-field toric code ground states under bit-flip and phase-flip decoherence. The authors show that stabilizer measurements act almost unitarily on undecohered ground states, and that the optimal post-measurement decoder can be implemented by Monte Carlo simulation of a constrained lattice gauge theory. Numerical results show a finite threshold close to the critical point, and for phase-flip noise the threshold is proven to be independent of the transverse field throughout the topological phase.
In briefQuantum Pointwise Dimension (QUPID): a method for spotting classical chaos in quantum systems
SummaryThe pointwise dimension measures the effective dimensionality of classical trajectories in phase space, with regular motion confined to a torus and chaotic motion filling an energy surface. This paper implements the measure in the quantum regime using Husimi representations of wavepackets, averaging them over long times to get a stationary phase-space density. The authors report excellent agreement with classical results for both autonomous and non-autonomous systems. They also explore obtaining the pointwise dimension from a time correlation function.
In briefBreaking quantum query barriers for graph collision
SummaryThe paper studies the graph collision problem, which asks whether a known graph on n vertices has an edge with both endpoints marked, accessed through queries to the marking input bits. The authors give a bounded-error quantum algorithm using Õ(n^{92/139}) queries, refuting a conjectured Ω(n^{2/3}) lower bound, and refine it to Õ(n^{11/17}) queries. They also prove an Ω(√(n log n)) quantum query lower bound, which exceeds the O(√n) certificate complexity barrier of the positive adversary method.
WhyThe authors refute the conjectured Ω(n^{2/3}) lower bound with quantum algorithms of Õ(n^{92/139}) and Õ(n^{11/17}) queries, plus a new lower bound.
Arjan Cornelissen, Amin Shiraz Gilani, Shiva Kintali et al.
In brief"Reading Why": An interpretable language model with Born-measured attention and exact word evidence, compiled to optics and qubits
SummaryThe paper builds a transformer in which every step is a Born measurement, with attention computed as |⟨q|k⟩|^4 normalised per row and a measured product replacing the feed-forward nonlinearity. On e-SNLI natural language inference it matches the test accuracy of a softmax transformer of the same size, and deleting the word its evidence ranks first changes the answer more often than for the softmax model. The same weights compile exactly to optoelectronic optics and qubits, and the attention part was partly run on an IBM quantum device.
In briefInterpretable quantum photonic generative models
SummaryThe paper proposes a physical generative model for correlated multivariate data, built on homodyne detection in photonic continuous-variable circuits. Two-mode squeezing provides correlations, local phase shifts tune their strength and sign, and an adjustable interferometer redistributes them across modes. Tested on standardized log returns of financial assets, the model reproduces rich correlation structures.
In briefA simple proof of when maximal entanglement is needed for local state verification and discrimination
SummaryThis paper studies when entangled resources suffice for verifying and discriminating quantum states using local (separable) measurements. Using only linear algebra, the authors give self-contained proofs of two known results: the Schmidt rank of the resource must be at least that of the state being verified, and if the ranks are equal, the resource must be maximally entangled. They then show that, among pure entangled resources of local dimension at most d, a maximally entangled state of C^d⊗C^d is necessary and sufficient for perfect local discrimination of a generic orthogonal basis, however weakly entangled the resource is.
In briefUniversal dynamics of quantum phase transitions with fractional long-range interactions
SummaryThis paper studies the full statistics of defects produced in a quantum phase transition within a nonintegrable long-range spin chain, the fractional Ising model. The authors use time-dependent variational-principle simulations based on matrix-product states and operators. The defect number cumulants follow power laws in driving time, with exponents that vary continuously with the fractional order. At low fractional order, defect decay is slower than the Kibble-Zurek prediction, while at intermediate order the mechanism holds.
Andra\'as Grabarits, Lincoln D. Carr, and Adolfo del Campo
In briefEffective Hamiltonians and non-Hermitian physics: compressing discarded-sector information into one boundary parameter
SummaryEvery realistic quantum system couples to an environment, and including it restores a closed Hermitian description, but the Hilbert space becomes intractably large. The paper uses the Feshbach–Löwdin projection to eliminate unwanted degrees of freedom exactly, yielding an effective Hamiltonian on the relevant subspace alone. This effective Hamiltonian generically becomes non-Hermitian when the eliminated subspace contains an accessible continuum and outgoing boundary conditions are imposed. Using a particle in a box with a delta barrier at its center, the authors reduce the discarded half-well to a single boundary parameter β, rederive the result via the Green's function and Dyson equation, and relate β to a self-energy term.
In briefDynaMite: A method that speeds up fault-tolerant quantum programs by adjusting measurement timing
SummaryFault-tolerant quantum computers spend most of their logical cycle time on qubit measurements, which take 43-86% of it. DynaMite is a lightweight real-time policy that selects the fastest measurement latency for each upcoming error-correction round, based on the remaining error budget and rounds left. Evaluations show it reduces logical cycle time by 1.69x on average (up to 2.29x) and lowers logical error rates to 0.46x of baseline on average (down to 0.13x).
In briefFast simulation of spectra and transport in multi-orbital quantum impurities
SummaryThis paper introduces C³L, a deterministic method combining coupled-Lindblad compact bath representations with coupled-cluster theory to compute steady states and Green's functions. It has polynomial computational cost and avoids the sign problem of quantum Monte Carlo. In a real-time multiorbital benchmark, it closely matches state-of-the-art quantum Monte Carlo while cutting computational cost by four to five orders of magnitude.
In briefQuantum query lower bounds via classical moment matching
SummaryThe paper develops a quantum version of the classical moment matching method, which is used to prove lower bounds in property testing. It presents two incomparable quantum moment matching theorems, with lower bounds of Ω(N^{1/2-1/(2r)}) and Ω(N^{1/2-1/(2(2^r-1))}), the latter holding for smaller parameter ranges. The proofs combine a new continuous hybrid argument with the polynomial method. The theorems are applied to obtain optimal lower bounds for k-distinctness, k-star-freeness, planarity and outerplanarity testing.
WhyThe authors introduce a quantum moment matching method combining continuous hybrid argument and polynomial method, yielding improved quantum query lower bounds for several property testing problems.
Chandrima Kayal, Fr\'ed\'eric Magniez, Sayantan Sen et al.
In briefComment on the paper "Supervised Learning with Quantum Hamiltonian Embeddings" (arXiv:2610.26833)
SummaryThis comment discusses how quantum Hamiltonian embeddings in arXiv:2610.26833 relate to earlier Quantum Cognition Machine Learning (QCML) work, including its prior implementation on IBM quantum hardware. It examines the novelty claims of the original paper. It also considers the computational role of encoding an entire dataset in a single ground state.
Alexander G. Abanov, Luca Candelori, and Kharen Musaelian
In briefData-Driven Ansatz Selection for Variational Quantum Circuits
SummaryThis paper gives three rules for choosing a variational quantum circuit's structure and depth from the data and gate set alone, before any training. The rules concern when entangling gates are needed, how the rank of the quantum Fisher information matrix grows with depth, and how many state directions Pauli-Z readout can resolve. On the Madelon dataset, the prescribed classifier has 239 parameters and reaches 78.5% test accuracy without a grid search. Ablations confirm the effects the rules predict.
In briefOperational thresholds for long-range entanglement in quantum networks
SummaryQuantum networks share entanglement between distant nodes by chaining short entangled links with local operations that succeed only probabilistically. The paper uses a stochastic extension of concurrence percolation on hierarchical (U,V)-flower networks to derive an exact recursion for end-to-end survival probability. For the (2,2)-flower it gives a closed-form phase boundary, and it finds a second threshold in initial link quality. The results show operational reliability and link quality can compensate each other.`
WhyThe authors incorporate probabilistic local operations into percolation theory, deriving an analytical phase boundary on hierarchical flower networks; results are reported only for specific network fa
In briefQECert: Lean Formal Foundations for Quantum Error-Correction Theory
SummaryQECert is a Lean library that provides reusable formal foundations for quantum error-correction theory. It connects error-correction criteria, stabilizer and CSS codes, and algebraic and homological constructions in one framework. Asymptotically good quantum LDPC codes serve as the main case study, with formalized constructions and asymptotic parameter theorems for hypergraph, balanced, and lifted products and quantum Tanner codes. The library also covers Bravyi-Poulin-Terhal bounds, constant-overhead magic state distillation, and energy barriers and storage errors for quantum memories.
In briefRevisiting the one clean qubit case of a semi-grand challenge
SummaryThis paper addresses Scott Aaronson's question of whether a quantum computer that always remains in a separable mixed state can be simulated classically in polynomial time. Earlier work claimed an affirmative answer for the one clean qubit model. The authors present a counterexample in which a pure clean qubit controls three gates acting on a maximally mixed three-qubit register, keeping all four qubits fully separable while producing a final unitary that does not match the form required by the earlier simulation result. They also give a continuous-time implementation and identify the gap in the earlier proof.
In briefWhether gravity-generated entanglement can be recovered when decoherence erases it
SummaryThe paper asks whether gravitationally generated entanglement that decoherence has made very small can still be recovered through strictly local quantum error correction. It analyzes Newtonian gravity with localization noise, including environmental noise and collapse-model noise from GRW, CSL, and Diósi–Penrose. For GRW and CSL noise, the authors show that local error correction can make the ratio of effective gravitational interaction strength to residual decoherence rate arbitrarily large. The Diósi–Penrose model does not show this enhancement under the bipartite protocol considered.
Yi-Xuan Wang, Julen S. Pedernales, Martin B. Plenio
In briefUltracold polar molecules as a platform for quantum science and technology
SummaryThis review article explains key concepts in the use of ultracold polar molecules for quantum science. Their intrinsic electric dipole moment couples strongly to microwave fields and can produce long-range, anisotropic, and tunable dipole-dipole interactions. The paper notes that such interactions can generate entanglement and enable study of quantum many-body phenomena. It also highlights recent experiments in optical tweezers and lattices and the cooling of molecular gases to quantum degeneracy.
In briefAssessing Quantum Computing-Ready Pipelines for Stochastic Unit Commitment in Power Systems
SummaryThis paper proposes a quantum computing-ready optimization pipeline for deterministic and stochastic unit commitment, combining classical, quantum-inspired, quantum annealing, and hybrid quantum-classical solvers on test systems including the IEEE 118-bus system. It introduces a QUBO reformulation and uses a divide-and-conquer strategy with Quantum Computing Monte Carlo simulation to handle 27,730 uncertainty scenarios. Results show the hybrid quantum method matches the classical baseline and cuts stochastic UC runtime from 40.76 to 14.79 hours, while pure quantum annealing is limited to small systems.
