Daily Summary for 2026-10-22

Introduction to the show: ident: Quantum Radio. Generated commentary on the latest quantum physics and condensed matter papers.

Kai: It's the twenty-second of October, twenty twenty-six, and this is the day's research.

Mira: 9 new papers came out today.

Kai: Hi, I'm Kai.

Mira: And I'm Mira, here with Lev.

Kai: We'll take the day in one pass, then pull out the papers we're staying with.

Mira: Start us off with Tensor computing application programming interface.

The summary: Kai: Welcome to October twenty second, twenty twenty six.

Mira: Today we focus on securing quantum machine learning pipelines against channel substitution attacks.

Lev: That is critical because customers cannot confirm the circuit they paid for is the one running in the cloud.

Kai: A malicious provider could swap a declared channel with one mimicking test results but differing fundamentally at the channel level.

Mira: The most significant work addresses this directly through QML-PipeGuard which detects these attacks while running by examining measurement statistics collected by the customer.

Lev: This detection method offers a two-sided guarantee.

Kai: If a substitution shifts a single-qubit output marginal by delta it forces a contract deviation of at least delta over the square root of three no matter how many qubits are involved.

Kai: This detection is supported by research on metric learning using quantum feature spaces.

Mira: Researchers introduced Quantum Projective Metric Learning or QPMeL to find exact closed-form kernels on unentangled manifolds.

Lev: This method avoids circuit simulation and solves gradient collapse by using a curriculum annealing approach that moves from a local observable regime to the global product kernel.

Kai: Additionally, there is foundational work on tensor networks aimed at solving portability issues through the Tensor Computing Application Programming Interface or TCAPI.

Mira: This lightweight API offers a single software interface for tensor network computations, allowing code written against it to move between different hardware and software platforms while keeping high performance.

Kai: So, the graph-theoretic formulation connects wavefunction signs to Max-Cut instances.

Mira: That suggests optimization relates directly to combinatorial problems.

Lev: Indeed, it shows how finding ground states is equivalent to solving weighted maximum cut problems on a graph structure.

Kai: That makes the connection between physics and discrete math clearer for us.

Mira: It helps frame the complexity of those systems in terms of established graph theory concepts.

Lev: And this relates directly to the work on Graph-Theoretic Formulation of Hamiltonian and Ground-State Sign Structures in Heisenberg Antiferromagnets.

Kai: That paper reveals a key equivalence for understanding antiferromagnet ground states.

Mira: It's interesting how that formulation translates optimization goals into graph problems.

Lev: Likewise, it provides a structural insight into the sign patterns of these systems.

Kai: We should look at QML-PipeGuard next; it deals with security in quantum pipelines.

Mira: That system detects when a malicious provider substitutes the quantum channel during QML computation by monitoring measurement statistics.

Lev: Monitoring statistics is a concrete way to identify those channel-substitution attacks.

Kai: It’s important for building trust in untrusted cloud environments for quantum tasks.

Mira: And then there is the Tensor computing application programming interface, which offers a unified and lightweight way to write portable and high-performance applications using tensor networks across different hardware.

Lev: That API simplifies writing applications involving tensor networks significantly.

Kai: It addresses portability, which is a major hurdle when dealing with diverse quantum hardware.

Mira: This makes complex computations more accessible across various platforms efficiently.

Lev: Furthermore, Metric Learning with Quantum Feature Spaces introduces QPMeL to learn metrics in quantum feature spaces by using an exact kernel that avoids circuit simulation and manages gradient collapse.

Kai: Avoiding circuit simulation is a big win for practical metric learning methods.

Mira: It handles gradient collapse effectively while still learning meaningful metrics in those quantum spaces.

Lev: That method offers a robust approach to learning useful representations in quantum feature spaces.

Lucky paper: 2512.23917: Kai: So we’re moving into something about making tensor network applications portable through a new interface, the Tensor Computing Application Programming Interface or TCI.

Mira: The core problem this paper tackles is that there isn't a unified software interface across tensor-computing frameworks, which really limits how portable these applications are.

Lev: This coupling means you’re tying algorithmic development directly to specific hardware and software back ends, which forces substantial code reimplementations when you try to migrate between different Tensor Computing Frameworks or TCFs.

Kai: The authors introduce TCI as an application-oriented, lightweight API designed specifically to solve this portability issue while maintaining high performance.

Mira: Architecturally, they sit between the tensor network applications and the different Tensor Computing Frameworks, using a well-defined type system built in C++seventeen.

Lev: That type system has three parts: an abstract tensor type called TenT, associated trait types exposed through tci::tensor traits, and a collection of auxiliary types.

Kai: And they’ve defined six categories of core tensor operations that cover everything from read-only queries to input output facilities.

Mira: Specifically, they list read-only queries like tci::order and tci::shape, along with routines for construction and destruction like tci::allocate and tci::zeros.

Lev: They also include operations for tensor manipulation that work on structure or elementwise values without needing to invoke tensor linear algebra primitives.

Kai: But they also cover the linear-algebra side, including contraction which reduces to ordinary matrix multiplication in some cases, and decomposition methods like Singular Value Decomposition or SVD.

Mira: The real selling point is the portability claim: numerical demonstrations show that codes written against TCI can be migrated across heterogeneous hardware and software platforms with performance comparable to native framework implementations.

Lev: They benchmark two specific applications, Application A computing the ground state of the one-dimensional transverse-field Ising model using iTEBD, and Application B simulating kicked Ising model dynamics on a heavy-hex lattice using a two-dimensional tensor network state combined with belief propagation.

Kai: So, even though they tested these specific things, the results confirm that TCI introduces only a negligible abstraction overhead when you compare it to native API implementations.

Mira: The future work mentioned involves maturing TCI into a fully standardized interface, planning asynchronous APIs for concurrent execution, and incorporating a standardized automatic differentiation interface in subsequent versions.

Lev: They also plan to extend TCI with a unified design that supports both dense and symmetry-structured tensors, aiming for migration to C++twenty or later standards eventually.

Kai: It sounds like the goal is to establish a viable pathway toward standardizing the software stack for generic, high-performance tensor computation on current and emerging HPC architectures.

Mira: It really moves the focus from just writing code for one framework to writing portable applications that benefit from a common interface, which is what TCI aims to do.

Lucky paper: 2605.00491: Kai: Let's get into this piece, Electrical detection of spin-flip transition in metal/heterostructure, which is paper number two of five on our list today.

Mira: This study looks at longitudinal magnetoresistance MR in thin metal films placed on an Ising-type antiferromagnetic insulator called Na5Co15 point 5Te6O36, or NCTO.

Lev: So the core observation is that the longitudinal MR shows steep changes with hysteresis during spin-flip transitions driven by magnetic fields along the easy axis of that NCTO crystal structure.

Kai: The researchers are targeting this antiferromagnetic insulator because it's considered a model material for an Ising antiferromagnet, and they found magnetic moments align along the c-axis due to the Ising anisotropy.

Mira: They noted that at zero field, near the Néel temperature of approximately fifty K, a partially disordered antiferromagnetic PDA state emerges where one of the Co(one) spins in each Co(one) triangle remains disordered.

Lev: As temperature drops further to about twenty K, those disordered moments freeze into a frozen PDA state called FPDA, and applying strong magnetic fields along the c-axis stabilizes induced ferromagnetic states with a sudden change accompanied by clear hysteresis.

Kai: The study investigated longitudinal MR measured under a magnetic field parallel to the electric current for metal/NCTO bilayers, and they saw similar MR jumps across Pt/NCTO, Pt/SiOx/NCTO, and Cu/NCTO heterostructures.

Mira: The origin of these MR anomalies is attributed to the magnetostriction of NCTO rather than just interface spin accumulation.

Lev: They estimated the magnetostriction coefficient lambda to be one point four times ten-four from data collected at ten K, which is more than an order of magnitude larger than values reported for NiO or Cr2O3.

Kai: That strain effect via magnetostriction means the dimensions of the metal/NCTO heterostructure can slightly change during the spin-flip transition, affecting the resistance of those metal films.

Mira: However, they observed a difference in MR spectra between Pt/NCTO and Cu/NCTO in the high-field regime where magnetic moments are fully aligned along the magnetic field direction.

Lev: The opposite signs of those slopes for Cu/NCTO and Pt/NCTO below fifty K suggest a weak antilocalization effect due to the strong spin-orbit interaction in the Pt films.

Kai: That's a key distinction, showing how the material of the metal layer matters in that high-field regime.

Mira: They also found that negative MR observed in Cu/NCTO below fifty K is likely ascribed to scattering caused by interface roughness.

Lev: Atomic force microscopy showed that the surface root-mean-square roughness of those as-grown NCTO crystals was at most four nm, which could contribute to the negative MR.

Kai: This suggests that roughness and interface scattering play a role in the intermediate temperature region for Cu/NCTO.

Mira: The study concluded that while high-field MR in Pt/NCTO is dominated by weak antilocalization at low temperatures, interface and surface roughness are important contributors to the negative MR in the intermediate temperature region.

Lev: So, this electrical detection of magnetic properties for an Ising antiferromagnet contributes directly toward developing antiferromagnetic spintronics applications.

Lucky paper: 2603.25742: Kai: We are looking at the pseudogap metal phase and non-Fermi-liquid criticality in bilayer nickelates today, which is paper number three of five for this show.

Mira: This study uses single-site dynamical mean-field theory, or DMFT, to investigate bilayer nickelates and finds a second Fermi liquid phase called the sFL in the underdoped regime.

Lev: The model uses the double Kondo lattice model with a ferromagnetic coupling JK equals negative two J H less zero and an antiferromagnetic exchange J sub t greater than zero.

Kai: So you're looking at how interlayer spin coupling directly dictates whether you get a conventional Fermi liquid or this second Fermi liquid phase.

Mira: The phase diagram maps the system as a function of J sub t, hole doping x, and Hund’s coupling JK equals negative two J H less zero.

Lev: The DMFT plus numerical renormalization group impurity solver calculations show a conventional Fermi liquid in the overdoped region with a Fermi surface area AFL equals one minus x over two.

Kai: But there is this second Fermi liquid phase, the sFL, in the underdoped regime with a much smaller Fermi surface area AsFL equals negative x over two.

Mira: The transition between these phases happens at a quantum critical point where the quasiparticle weight Z vanishes and there is a jump in delta A F S of about one half.

Lev: In that critical regime, we see non-Fermi liquid behaviors, specifically a log divergence of susceptibility for bosonic operators like the total density and layer-alternating spin.

Kai: The pseudogap behavior in the sFL phase is shown as a divergent structure in the self-energy approximated by sigma omega equals phi two over omega plus mu psi.

Mira: This mathematical form leads to a dip in the spectral function Ac equals omega at positive energy, and an emergent composite fermion dominates at low energy.

Lev: The ancilla framework gives us a model where the pseudogap comes from hybridization between an electron and an ancilla fermion, creating a small hole-like Fermi pocket around pi pi.

Kai: That means we can picture the pseudogap arising from this specific hybridization in the wavefunction.

Mira: The interlayer spin coupling J sub t plays a crucial role in stabilizing that sFL phase; for strong ferromagnetic Hund’s coupling JH greater than or equal to two point five t, the sFL is stabilized by stronger J sub t and smaller doping x.

Lev: The critical doping xc increases as the interlayer spin coupling J sub t increases in that regime.

Kai: Conversely, when we have anti-Hund’s coupling JK greater than zero, a larger J sub t favors the conventional Fermi liquid phase through an RKKY-like mechanism.

Mira: This suggests that interlayer singlet formation competes with Kondo hybridization when anti-Hund’s coupling is present.

Lev: Finite interlayer hopping t sub t breaks the separate U one t times U one b symmetry down to a single global U one symmetry, which is important for understanding the stability of non-Fermi liquid behavior near that quantum critical point.

Kai: The results suggest that the strange-metal behavior we see in bilayer nickelates might originate from this quantum critical point between the pseudogap metal and the Fermi liquid.

Mira: There is also a deep connection suggested between bilayer nickelates and twisted bilayer graphene, as a similar sFL phase may control the normal state of twisted bilayer graphene.

Lev: The paper connects that quantum critical point to the non-Fermi liquid QCP of the two-impurity Anderson model with a similar structure.

Kai: The ancilla fermion can be interpreted as a spin-polaron like trion operator, which is identified through an equation-of-motion hierarchy of the electron Green’s function.

Mira: That composite operator carries the same charge, spin and fermionic statistics as the original electron operator.

Lev: A variational wave function can be constructed by projecting that ancilla wave function onto the physical Hilbert space, which is expected to capture the correct low-energy physics of those systems.

Kai: While direct experimental detection of those small Fermi pockets is hard, their presence is inferred from transport measurements.

Mira: In the sFL phase, spectral weight on the original dispersive band equals negative one over pi times imaginary alpha omega is strongly suppressed, and most of it moves to a nearly flat band at low energy.

Lev: That flat band corresponds to the contribution from psi j i alpha together with a broad high-energy feature associated with psi u i alpha.

Kai: The evolution of that small Fermi surface gradually broadens into a large Fermi surface-like structure at energies of order one over two t.

Mira: The study successfully identifies both a conventional FL in the overdoped region and this pseudogap metal sFL in the underdoped regime, separated by that non-Fermi liquid critical region.

Lucky paper: 2312.01655: Kai: We’re moving on to our fourth paper of five today, and it’s titled Metric Learning with Quantum Feature Spaces.

Mira: This paper proposes a quantum-aware method for classically training dense and high-quality quantum encodings using something called QPMeL.

Lev: The core idea is mapping classical data points onto the surface of independent unit spheres in R three which naturally aligns with the state space of multiple unentangled qubits.

Kai: So, they define polar and azimuthal angles, theta and gamma, from the classical data that map directly to quantum states using only RY and RZ gates.

Mira: They then introduce a Projective Metric Function or PMeF which computes similarity in quantum state space using just those R three coordinates.

Lev: This PMeF is constructed by combining polar-to-cartesian conversion with a Complex Kernel Function to handle the coordinate periodicity and complex values.

Kai: During training, QPMeL runs entirely in classical real-valued space, using that PMeF for metric loss functions like Prototypical loss.

Mira: This is important because it avoids barren plateaus by training classically and sidesteps the challenges of training actual quantum circuits.

Lev: They also introduced a gradient trick for the PMeF, modifying the metric used during training by replacing the product with a sum over single-qubit fidelities to stabilize those gradients.

Kai: That modification helps make sure the models converge more consistently during training.

Mira: The performance numbers are pretty strong here, especially for few-shot learning; they achieve ninety percent accuracy on the train set even in a fifteen-way-one-shot setting using only twenty qubits.

Lev: They also show state-of-the-art results on MNIST and FashionMNIST, achieving ninety-nine percent accuracy on binary tasks for those datasets.

Kai: That’s good scaling; it’s the only method they showed that scales successfully to a ten-class classification on both of those datasets while hitting around eighty-five percent accuracy on FashionMNIST.

Mira: And the paper points out that while it might not learn entangled states, the results show this doesn't seem to affect performance at all.

Lev: The resulting quantum state can then be used as input for downstream applications like generative or regressive models, and you only need a single RY and RZ gate per qubit for encoding.

Kai: That means you can use any circuit in place of static angle encoding because the downstream application circuit can be trained separately without needing to retrain QPMeL.

Mira: The limitation they state is that it doesn't learn entangled states, but they argue that this doesn't seem to hurt the performance metrics they reported.

Lev: In conclusion, QPMeL introduces a framework that learns encodings classically via spherical surfaces and a PMeF, proving that classical training can yield dense embeddings respecting Hilbert space geometry while avoiding quantum circuit training instability.

Kai: So, we have this solid classical approach for learning quantum representations efficiently for NISQ devices.

Mira: It shows how we can get high accuracy with very few qubits when using this QML approach.

Lev: It’s a powerful tool that bridges the gap between classical data and quantum state space geometry without needing complex circuit training overhead.

Lucky paper: 2209.09852: Kai: We're moving into our fifth and final paper of the day, a deep dive into the evidence for non-unitary triplet-pairing superconductivity in noncentrosymmetric TaRuSi and comparison with isostructural TaReSi.

Mira: This study uses muon spin rotation and relaxation along with density functional theory calculations to see what kind of pairing symmetry these materials have.

Lev: The transverse field muon spin rotation measurements show that TaReSi exhibits isotropic s-wave superconductivity, but TaRuSi shows multi-gap superconductivity.

Kai: That multi-gap nature in TaRuSi is interesting because it suggests something more complex than a simple single gap structure.

Mira: The zero-field muon spin rotation measurements on TaRuSi reveal small spontaneous fields that show an onset when superconductivity kicks in, which implies broken time-reversal symmetry.

Lev: That broken time-reversal symmetry is what the authors link to a non-unitary triplet pairing state, and they attribute this to the relative weakness of spin-orbit coupling in TaRuSi compared to its strength in TaReSi.

Kai: So, while both materials are unconventional, the specific type of pairing seems dictated by how much spin-orbit coupling is present.

Mira: The paper discusses how noncentrosymmetric superconductors allow for a mixture of spin-singlet and spintriplet Cooper pairs when antisymmetric spin-orbit coupling is allowed by symmetry.

Lev: They find that the electronic antisymmetric spin-orbit coupling is forty-one meV for TaRuSi and eighty-one meV for TaReSi, which explains why the effect on the band structure is relatively weak in TaRuSi compared to other noncentrosymmetric superconductors.

Kai: That comparison between those two spin-orbit coupling values helps explain the difference in their superconducting symmetries.

Mira: The paper also looks at how electron-phonon coupling strength is estimated by matching the computed density of states at the Fermi energy, D(EF), to values like two point two eight and three point three four states eV f.u. for TaReSi and TaRuSi, respectively.

Lev: Both of those values are close to what you would expect from the McMillan formula, which points toward moderate electron-phonon coupling in both materials.

Kai: It seems they’ve established that the underlying crystal structure plays a major role in determining these allowed pairing symmetries.

Mira: They also use the Uemura classification to place TaRuSi and TaReSi near other unconventional superconductors like cuprates and iron-based superconductors, showing their Tc over magnetic field ratios are in the range of zero point zero one to zero point one for unconventional materials.

Lev: And that ratio is much smaller than what you see in conventional superconductors, which sets a clear benchmark for what counts as unconventional behavior here.

Kai: The study concludes that TaRuSi is the first system other than LaNiC2 with the C2v point group to host such a non-unitary triplet-pairing state.

Mira: That means they’ve found an analogous system in TaRuSi that shares the same symmetry breaking as LaNiC2, despite having different structural elements like Re versus Ru.

Lev: The authors suspect that Hund’s coupling between electrons on Ni orbitals could be the pairing mechanism for non-unitary triplet pairing in TaRuSi, similar to what's seen in LaNiC2.

Kai: So, the link between that specific orbital coupling and the resulting superconducting state is a key hypothesis they propose.

Mira: It’s a big piece of evidence suggesting that structural differences like those between TaReSi and TaRuSi directly manifest as differences in superconducting symmetry.