Quantum papers — 2026-10-20

Today's focus is on how we can get clearer pictures in noisy quantum settings, which is crucial because reliable characterization is the backbone for building scalable quantum computers. We are looking at methods to estimate the fidelity of noisy devices without needing a full state tomography, as standard techniques become impractical with larger systems.

One approach involves direct fidelity estimation where we use importance-sampled Pauli measurements to estimate the fidelity between a target state and an experimentally prepared one. We found that adding attenuation-aware correction improves robustness against measurement noise.

Another method is based on survival probability, which reconstructs fidelity from accessible moments of noisy evolution and is intrinsically robust against preparation-and-measurement errors. In this case, an L two optimized expansion substantially outperforms naive Taylor truncation in high-noise regimes.

Beyond state characterization, we are exploring ways to manipulate quantum evolution itself through periodic conditional measurements on photonic chips. This work shows that by repeatedly interrupting coherent evolution with local measurements, we can engineer conditional quantum states and manipulate detection statistics using bright and dark modes of the conditional evolution operator.

This leads into how we can control these dynamics, as seen in time-optimal robust quantum control for partially known errors. This suggests that knowing the error class shortens the minimum time required for optimal control.

Furthermore, we are looking at practical ways to build and read out complex quantum circuits efficiently. Patchwork introduces a lattice-surgery compiler that automates multi-qubit Pauli measurements on surface-code layouts. It uses geometric rules to simplify stabilizer placement and scheduling, which helps achieve total failure probabilities close to monolithic decoding for sequential measurements.

This contrasts with efforts in connectivity bounds for GHZ graphs, where we prove upper bounds on the dimension of multipartite entangled states based on graph connectivity. This sets limits on what kind of high-dimensional states we can realistically create.

Finally, when considering the physical realization of these computations, we look at the energetics of fault-tolerant spin-qubit quantum computation. This work benchmarks a Loss-DiVincenzo spin-qubit computer and shows that optimizing control signal demultiplexing can reduce the overall energy cost by about two orders of magnitude. This suggests quantum computing could become energetically competitive before achieving a speedup in execution time at scale.

The work concerning rectification mechanisms in correlated quantum systems is particularly important because it provides a pathway for controlling transport in technologies that rely on coherent quantum circuits. This research establishes a mechanism for both particle and energy flow in fermionic systems coupled to external baths, even when there is no bias between those baths, by using a time-periodic drive modulated locally by a tilted potential.

These drive-induced currents are found to be enhanced at specific resonant frequencies where avoided crossings occur in the Floquet quasienergy spectrum. This means the Floquet states act as important intermediate channels for transport.

The study then moves to how introducing a bias between the baths interacts with this drive-induced current, showing that it can either reinforce, compete with, or even cancel the bias-induced contribution. This interplay allows for rectification in the non-interacting regime, which is quantified by a coefficient R that measures asymmetry between transport up and down the tilt.

Furthermore, they identify a regime of bipolar rectification where current flows only up the tilt regardless of bias polarity. Perfect rectification is achieved at the boundary between standard and bipolar regimes when reverse-bias current is completely suppressed.

In terms of complexity, when interactions are included between particles, this region supporting bipolar rectification expands significantly and R can be enhanced by many orders of magnitude. This mechanism for particle and energy transport has potential for experimental realization in condensed matter and cold atom systems.

This contrasts with work on qubit readout errors, where the Readout Error Decoder (RED) converts repeated thresholded binary records into a channel-resolved error budget. This identifies limiting error channels without needing to discriminate non-computational states.

The study on minimal models for Weyl nodes and Fermi arcs in PtBi 2 provides a microscopic foundation for understanding surface superconductivity by developing a model guided by density functional theory calculations. This model captures the evolution of Weyl nodes with spin-orbit coupling, allowing it to describe the spin-momentum locking of surface Fermi arcs.

This topological insight connects to research on surface-extended Andreev bound states emerging from intrinsic surface superconductivity in PtBi 2. This uses scanning tunneling microscopy to reveal a spatially homogeneous gap and detects surface-extended in-gap states interpreted as Andreev bound states.

Finally, the work on quantifiers of noise reducibility under restricted control reinterprets previous numerical findings on dynamical decoupling and non-Markovian memory by showing that the conclusion about dynamical decoupling as a resource distillation still holds. This holds even when using new quantifiers inspired by quantum combs.

This connects back to simulation methods like NESSi 2.0, which extends nonequilibrium Green's functions by truncating memory integrals to reduce computational cost while introducing functionalities for describing nonequilibrium steady states relevant for transport settings.

The most significant work here is the development of efficient LLM-generated shuttling compilers for complex trapped-ion architectures because it drastically cuts down the time needed to design new quantum hardware control software. This moves development from months down to days.

This method involves using a frontier large language model, specifically Claude Opus 4.7 and later Claude Fable 5, to generate and refine Python code that translates high-level specifications into sequences of ion movements for shuttling compilers AKMPC. This approach has shown remarkable success, reducing the number of shuttling timesteps by up to seventy-two percent for linear segmented traps and forty-three percent for junction traps AKMPC.

The ability of these LLMs to produce working, correct compilers without extensive manual algorithmic engineering is a major breakthrough. This contrasts with traditional hand-crafted methods, which the LLM results often surpass on larger circuits AKMPC.

For densely connected architectures, the required shuttling timesteps can drop by an order of magnitude compared to corridor-like structures AKMPC. This efficiency gain is important because it means that the complexity of the underlying ion trap design is less penalized in terms of control sequence length.

Another area showing promise is in simulating quantum dynamics, where Sparse-QITE introduces a regularized formulation to create compact generators for quantum imaginary-time evolution Sparse-QITE. This technique aims to reduce the number of quantum operations needed for QITE by promoting sparsity.

Showing that in some high-accuracy regimes, it requires only about one percent of the standard QITE gate count Sparse-QITE.

In terms of fundamental physics, the work on symmetry-adapted eigensolvers addresses how to efficiently find ground states in strongly correlated systems by separating where the wave function is sampled from where it is defined Symmetry-Adapted Eigensolver. This method uses a real-space amplitude to generate configurations directly, which ensures that every generated configuration is an unbiased sample of the desired state Symmetry-Adapted Eigensolver.

Finally, there is research into measurement dependence in quantum correlations. A paper establishes tight lower bounds on the minimal measurement dependence required for local deterministic models to reproduce Bell inequality violations Exact minimum measurement dependence for faithful local deterministic models of multipartite GHZ-Mermin correlations. This work provides rigorous mathematical constraints on how much a hidden state must depend on the measurement settings to achieve specific quantum correlations Exact minimum measurement dependence for faithful local deterministic models of multipartite GHZ-Mermin correlations.

The work on programmable all-to-all entangling gates for cavity-coupled atoms is most significant because it directly addresses a major hurdle in building fault tolerant quantum computers: achieving both high fidelity entangling operations and non-local qubit connectivity. This research tackles the problem of local entanglement, where separated atoms need to be physically moved, which introduces motional errors.

The authors design a Mølmer Sørensen-like gate between neutral atoms coupled to a photonic cavity, using the shared cavity coupling to enable all-to-all connectivity without moving the atoms themselves.

This approach is powerful because it makes the gate robust against changes in the number of photons in the cavity and allows for selective entanglement between any subset of separated atoms using local laser addressing. This eliminates motional errors and erasure, leading to demonstrated gate fidelities exceeding zero point nine nine, with higher fidelity possible through optimal control of pump amplitudes.

This capability is complemented by the work on chiral valence-bond solids which shows that spin-singlet bond order can spontaneously acquire handedness without magnetic order.

This spontaneous chirality in a chiral valence-bond solid is inherited by chiral triplons, whose spin textures reverse between two enantiomers, generating a temperature-gradient-induced net spin response. This allows for the direct detection of chiral triplons and the selected handedness via polarized inelastic neutron scattering. This establishes a link between spontaneous chirality and magnetic excitations.

This connects to the understanding of magnetic structures in organic materials where C 60 molecular films modify buried spin textures by enhancing effective magnetic anisotropy and interfacial Dzyaloshinskii-Moriya interaction.

Furthermore, research into superconducting circuits shows that multilayer cuprates like Bi two two two three exhibit an unexpected simultaneous enhancement of both pairing strength and phase coherence when interfacing planes with distinct hole concentrations. This unique interlayer mechanism provides new microscopic insight into the record critical temperatures achieved in these materials.

This contrasts with the work on 3D-integrated superconducting quantum circuits, which shows that using an asymmetric indium architecture combined with multi angle compression can preserve qubit coherence times comparable to planar devices through low resistance interconnects.

The work concerning exact quantum splitting and the structure of finite algebras is particularly important because it provides a deterministic way to perform factorization in circuit models. This has direct implications for certain types of quantum computation.

This method allows for an exactly computable single-qubit rotation that depends on the input, and it shows that factorization takes exactly r-1 splitting rounds with O(n cubed q) quantum operations when combined with classical preprocessing AKMPC.

Furthermore, this technique can extract primitive idempotents from arbitrary finite-dimensional commutative unital algebras using O(n 4 q) quantum operations. This is a significant structural result for these algebras.

The development of quantum conjugate gradient descent is also noteworthy because it offers a geometrically informed approach to ground-state preparation that goes beyond standard steepest descent methods. This Quantum Conjugate Gradient Descent algorithm achieves faster convergence per iteration than the baseline double-bracket quantum imaginary-time evolution.

Although this speed comes with a trade-off in circuit depth for each update step Quantum-compatible Riemannian conjugate gradient for ground-state preparation, this geometric framework is complemented by the study of quantum geometry of connected state manifolds. This shows how diabolic points can act as bridges between eigenstate manifolds, creating a single connected state manifold Quantum geometry of connected state manifolds: When diabolic points act as bridges between eigenstate manifolds.

Another area of progress involves understanding the dynamics in open quantum systems, specifically studying semiclassical Ehrenfest paths where the Fokker-Planck equation governs the mixing measure Semiclassical Ehrenfest paths in open quantum systems. This work provides a phase-space interpretation of how classical trajectories emerge from open quantum dynamics in the frictionless harmonic regime.

These dynamical insights are further complicated by noise, which is addressed by studying magic in quantum trajectory unravelings of noisy random quantum circuits Magic in quantum trajectory unravelings of noisy random quantum circuits. This research shows that an optimal choice of Kraus operators can substantially reduce the extensive magic density of trajectories without changing the linear scaling with system size.

Finally, there is work on reducing complexity for practical applications, such as drone delivery planning. Classical feasibility preprocessing significantly reduces the search space before applying QUBO-based optimization Reducing QUBO Complexity in Drone Delivery Through Classical Feasibility Preprocessing. This hybrid approach demonstrates that reducing the binary search space through classical checks yields substantial improvements in runtime for solving the resulting quadratic unconstrained binary optimization problems.

The work by Li and Li on Multiplicative Semidefinite Programming Upper Bounds on Quantum No-Signalling-Assisted Zero-Error Capacity is particularly important because it tightens the known upper bounds for quantum no-signalling-assisted zero-error capacity using a semidefinite programming parameter Gamma(K). They show that this parameter is the tightest among those considered.

This work matters because it provides sharp bounds for classical and quantum channels, showing that the capacity is at most Gamma(K) or (1 plus r 1/2)Theta NS/2 in certain cases, and it resolves an open question about the asymptotic simulation cost S 0, NS(K).

The study by Gao et al. on Thermodynamic-limit bootstrap for solving fractional quantum Hall ground states is significant because it develops a quantum many-body bootstrap method that solves these ground states directly in the thermodynamic limit without relying on finite system extrapolations. This method yields rigorous and quantitatively tight lower bounds on the energy and identifies the topological order, successfully resolving the ground states at nu equals one over four and twelve over five.

This approach is applicable to both Abelian and non-Abelian quantum Hall states.

The paper by Tsunoda et al. on localized states of the non-Hermitian Schrödinger equation with a complex one over r squared potential is noteworthy for bridging Hermitian few-body physics and non-Hermitian many-body physics. It shows how Efimov bound states smoothly connect to bipolaron states when the imaginary part of the potential is increased.

They found that each bipolaron state is accompanied by another localized state that oscillates on the scale of the mesh size, which has no counterpart in the continuum limit.

Shalby et al.'s work on Stitched surface codes presents a fault-tolerant quantum-memory architecture that concatenates four distance-t rotated surface-code patches with an outer code. This achieves logical error rates scaling as p t. Increasing the code distance from d equals four to d equals six reduces the block logical error rate by a factor of 1.52.

The paper by Ghosh and Singh on Local finite rate semi device independent randomness expansion shows that contextual correlations with only a single localized party are sufficient for secure randomness expansion in a semi-device independent manner. This suggests that recent experimental tests of non-contextuality inequalities can be modified to achieve secure randomness expansion following their proposed protocol.

Finally, the work by Defant Agustina G. Magnoni et al. on Loss-independent time-delay estimation via Hong-Ou-Mandel photon bunching detection demonstrates that using a four-detector scheme achieves the maximum precision given by the quantum Fisher information without needing to characterize system losses. This four-detector scheme outperforms the two-detector scheme in precision for time delay estimation, improving dispersion by a factor of two for repetitions.

The work on bubble-sort exchange networks matters because it shows how to use simple exchange interactions in quantum dots to perform complex computations like language model distillation. This demonstrates that routing can be achieved purely through entanglement, not just by using other types of pulses, which is a key insight for building scalable quantum circuits.

The core idea involves building a variational bubble-sort exchange network where the odd-even transposition is done with one parameter per pulse AKMPC. This network conserves total spin, which means the reachable readout distributions are strictly limited by the initial state's total spin content.

The selection rule derived for the two-spin singlet-triplet encoding states that outcomes with k triplet pairs are bounded by the initial weight in sectors S <= k, excluding S=0 when k is one. This rule is sufficient because a transportation argument proves that the reachable distributions form the polytope defined by this rule, and its total-variation floor is reached.

This constraint contrasts sharply with other states; for instance, products of singlets and T0's forbid entire outcome classes, while the Neel state forbids none, and the three-spin EO code has no floor at all. This limitation sets a boundary on what can be achieved in this specific network architecture.

The distillation process involves using matrix-product-state Born machines as classical control to distill next-token distributions from GPT-2 over restricted vocabularies into the network structure of AKMPC. The results show that at six logical qubits, the three-spin chain achieves a KL divergence of 0.011 nat at depth N and 0.004 at 2N, which is below the bond-dimension-four MPS limit; this is compared to the two-spin chain reaching 0.28 to 1.5 from singlet/triplet products for the first teacher AKMPC.

The work on efficient quantum simulation of unitary gates based on atomic ensembles provides a different route to dynamics, using the Cartan decomposition method to construct symmetry-adapted generators for implementing ensemble Hamiltonian evolution Efficient quantum simulation of unitary gates based on atomic ensembles. This approach is efficient because the Cartan protocol reproduces operator dynamics with zero error that is substantially smaller than Trotterization.

The error remains independent of atom number while Trotter error increases with system size Efficient quantum simulation of unitary gates based on atomic ensembles.

The concept of circuit duality in gyrators for superconducting circuits suggests a powerful transformation where non-reciprocal circuits can be mapped to dynamically equivalent reciprocal ones Gyrators for superconducting circuit design. This formalism extends Tellegen's result, showing that gyrators make reciprocal elements behave like their duals in non-linear contexts like Josephson junctions.

The path-degenerate quantum interferometry scheme aims to enhance gravitational-wave detector sensitivity by reducing spatial mode mismatches without needing extensive optical subsystems Path-Degenerate Quantum Interferometry for Loss Resilience in Gravitational-Wave Detectors. This scheme experimentally achieved a shot-noise-preserving signal enhancement by consolidating optical components, leading to a 110 kbps secure key rate with a QBER of 3.28 percent under active beam stabilization.

Finally, the study on quantized transport in Floquet topological insulators shows that two-terminal and Hall conductance are both quantized as W epsilon e 2/h when summed over all Floquet sidebands Quantized Transport in Floquet Topological Insulators. This quantization is achieved only after summing the contribution of all sidebands, and the convergence is fast enough to make observation accessible in many parameter regimes.

The most significant development concerns how quantum algorithms can tackle hard combinatorial optimization problems, specifically integer programming. The work by Qudit-based scalable quantum algorithm for solving the integer programming problem Qudit-based scalable quantum algorithm for solving the integer programming problem shows a quantum speedup over classical methods for general polynomial IP problems. This achieves a reduction of about d n/2 in time complexity compared to classical brute force.

This is important because IP is notoriously difficult for both classical and existing quantum approaches that use qubits.

This improvement builds on earlier work addressing resource inefficiency in mapping integer variables to qudits Qudit-based scalable quantum algorithm for solving the integer programming problem, which used qudits to address this issue but lacked scalability.

That work is complemented by the exploration of random quantum circuits and monitored dynamics in Les Houches lectures on random quantum circuits and monitored quantum dynamics, which applies statistical mechanics to study these complex dynamics where exact descriptions are often intractable.

In terms of fundamental encoding techniques, the work on Quantum block encoding for one-pair semiseparable matrices Quantum block encoding for one-pair semiseparable matrices provides a new approach for embedding these specific rank-structured matrices into larger unitary forms. This requires only O(polylog (N)) time under certain oracle assumptions.

This technique is less explored than methods focused on sparse matrices, which is why this work on one-pair semiseparable structures matters.

Other findings touch upon physical systems; for instance, the simulation of photo-doped Mott insulators Numerically exact simulation of photo-doped Mott insulators uses Quantum Monte Carlo to show that the Mott gap remains robust even with large photo-doping. This reveals strongly renormalized quasiparticles.

This contrasts with studies on electron interactions in interferometers, where time-resolved simulations suggest that while electron-electron interactions renormalize pulse velocity, the interference effects remain robust Effect of electron-electron interactions on the propagation of ultrashort voltage pulses in a Mach-Zehnder interferometer.

Finally, theoretical physics explores quantum correlations in photon systems through multiband dispersion Multiband dispersion and warped vortices of strongly-interacting photons. This describes how Rydberg polaritons exhibit warped band structures due to strong interactions.

This physical modeling is distinct from the computational work on quantum hardware, such as probing deconfined quantum criticality on a digital quantum computer Probing Deconfined Quantum Criticality Point on a Digital Quantum Computer. This demonstrates that Landau-forbidden transitions can be resolved using current hardware by distinguishing phases through order parameters.

Today's papers

The papers

Important terms

Fidelity Estimation
Methods to estimate how accurate a noisy quantum device is without needing a full state tomography. This includes using importance-sampled Pauli measurements and attenuation-aware corrections for better noise robustness.
Survival Probability
A technique that reconstructs fidelity from accessible moments of noisy evolution. It's robust against preparation and measurement errors, especially when using optimized expansions over naive truncation.
Conditional Quantum States
Engineering quantum states by repeatedly interrupting coherent evolution with local measurements on photonic chips. This allows for manipulating detection statistics using bright and dark modes.
LLM-generated Shuttling Compilers
Using large language models like Claude to automatically generate and refine Python code for ion trap shuttling. This drastically cuts down the time needed to design complex quantum hardware control software.