Papers

The objective of this paper is to demonstrate the future impact of a dedicated low-earth-orbit (LEO) Global Navigation Satellite System (GNSS) on positioning, navigation, and timing (PNT), in addition to a reflective discussion on how common GNSS evaluation methods can best reflect LEO satellite dynamics. This paper will feature Xona's LEO constellation, Pulsar, at various stages of deployment in a case study comparing GPS, LEO-augmented GPS, and standalone LEO. Results include visibility and dilution of precision (DOP) analysis for Pulsar at four stages of constellation deployment, as well as two cases exploring how precise point positioning (PPP) convergence is affected by Pulsar's LEO satellite dynamics. Pulsar is highlighted in this paper for its purposeful design that allows for both GNSS augmentation through compatible signals and GNSS independence (Gala et al., 2026). As dedicated LEO satellite-based positioning becomes more prominent in the PNT industry, it is important to consider which methods are used to evaluate constellation geometry and positioning performance. By evaluating the constellation through a variety of different methods in this paper, a discussion on the future evolution of LEO navigation constellations is offered. Pulsar is compared to MEO GNSS, both as an augmentation and a standalone service, through availability, dilution of precision, and PPP convergence.

We study collective state preparation and channel encoding in fermionic systems under parity superselection and composition governed by the canonical anticommutation relations. For a two-mode Gaussian state $\tau$, we obtain the exact entanglement of formation, $E_{\mathrm F}^{\mathrm{phys}}(\tau^{\otimes m})=\lceil m/2\rceil\log2$ for every $m\geq1$. Two copies have the same formation value as one, violating additivity and strong superadditivity, while regularization halves the single-copy value. For an associated Gaussian channel, we determine the minimum output entropy and Holevo information at every blocklength and prove nonadditivity of both. Together, these results give exact answers to the fermionic counterparts of the four additivity questions associated with Shor and Hastings. One physical use carries no classical information, whereas two uses transmit one bit perfectly. The unassisted classical and zero-error classical capacities both equal one half bit per mode. A sharp finite-block bound gives a strong converse: the decoding success probability vanishes at every rate above capacity. These optimizations cover all physical pure-state decompositions and channel inputs, with Gaussian ensembles and classical codes attaining the optima. The formation and minimum output entropy formulas hold for every Rényi order $\alpha\in(0,\infty]$. We also determine the unassisted quantum and zero-error quantum capacities, both equal to one half logical qubit per mode, and prove a strong converse: entanglement fidelity vanishes at every rate above capacity. For a noisy Gaussian family, we determine the one- and two-copy formation values and prove strict nonadditivity at full rank. For $\tau$, Bell violation under local parity-preserving measurements first appears at three copies.

We study continuous-time supercritical branching random walks in space-inhomogeneous random environment on the Hamming cube $\{0,1,\dots,b-1\}^d$, where the reproduction laws at each site are i.i.d. sampled. This serves as an idealized model for RNA sequence evolution with mutation–selection balance. The reproduction and mutation events are decoupled, and we assume that the reproduction law has essential supremum strictly below the deterministic mutation rate. Our main results provide tight asymptotics of the first-passage times for the model, uniformly in the origin–target distance $1\le m\le d$ as $d\to\infty$, conditional upon survival. We prove both quenched tightness in environment probability and annealed tightness around a deterministic center. Moreover, we identify the leading order of the deterministic center and develop an expansion in the sparse-distance regime $m=o(d)$. Our proof technique derives quantitative approximations of the model using an independent-visit variant, where revisits still resample the environment. As an application, we show that on a macroscopic scale, increasing the reproduction law in convex order decreases the first-passage times, and increasing the mutation rate increases the first-passage times for the sparse regime.

The 248 keV nuclear-recoil event observed by the LUX-ZEPLIN (LZ) experiment could potentially arise from dark matter (DM) interactions with nucleons. We investigate elastic scattering interpretation, focusing on Lagrangians that yield a local significance of $3\sigma$ or higher for some DM mass. We identify DM parameter space compatible with the LZ event and assess its consistency with collider constraints by recasting the ATLAS monojet plus missing-transverse-momentum search at $\sqrt{s}=13$ TeV, based on $139~\mathrm{fb}^{-1}$ of data. For most of the effective interactions, we find that LHC constraints exclude the DM interpretation of the LZ event for DM masses below $\lesssim\,$TeV subject to the effective field theory (EFT) validity. In particular, we derive, for the first time, LHC constraints on several explicit momentum-dependent Lagrangians, including $\mathcal{L}_{10}$. For many interactions, these constraints are sufficiently strong that the corresponding parameter space below DM mass $\lesssim\,$TeV may be challenging to probe even with future direct detection experiments. Our results, therefore, have broader implications beyond the interpretation of the LZ event itself.

Self-testing certifies quantum states and measurements from observed correlations alone, without requiring a description of the devices that produce them. In this work we establish, to the best of our knowledge, the first robust self-test of the full continuous family of measurements generated by Majorana operators. We achieve this by showing that the Grothendieck game, a two-player nonlocal game in which the players receive directions on a sphere and return binary answers, is a self-test of Majorana observables. That is, any near-optimal performance certifies an entire continuously parameterized family of quantum measurements. We further extend this certification to products of Majorana observables, obtaining a robust self-test of the spin representation of the continuous pin group.

We study the emergent randomness in projected ensembles generated by dual-unitary circuits. Specifically, we consider evolution by solvable dual-unitary brickwork circuits followed by measurements on part of the system, which induce an ensemble of states on the unmeasured subsystem. Previous work has established that this projected ensemble of states forms a quantum state design in the limit where the bath size is large, but the rate of convergence at finite bath size has not been rigorously established. Under three inverse-polynomial assumptions, local permutation mixing, a bound on recurrent backward motion, and contraction at the measured boundary, we fill this gap and prove that a bath of polynomial size suffices to form approximate $k$-designs. In establishing our results, we develop a microscopic transport interpretation of emergent randomness in the projected ensemble, in which deviations from local permutation operators appear as defects, and make connections with the theory of quantum walks.

Decoded quantum interferometry (DQI) uses classical error correction to solve optimization problems, but it is hard to tell which problems could give it a quantum advantage. We develop soundness rules that screen candidate problems by their code parameters, decoding guarantees and classical baselines. For the standard objective of counting satisfied constraints, we prove that under worst-case Hamming decoding DQI's guarantee can exceed Prange completion by at most $(\sqrt{2}-1)/2$. Optimal polynomial intersection (OPI) attains this margin cap asymptotically, so no candidate in this setting can exceed OPI's margin. We also derive a performance law for real-valued objectives: the objective fixes both the solution quality and which errors the decoder must correct. An AI-agent search guided by these rules found two candidate applications. Alternant max-agreement on McEliece-type public keys tests what a secret decoder is worth to DQI against classical solvers that see only the public instance. Multiplicative polynomial intersection, a knapsack-type problem built from discrete logarithms, pairs a cosine objective with a public decoder for signed errors. For both we prove decoding guarantees and finite-length comparisons with Prange restarts, and on the larger instances we tested, DQI's expected quality exceeds what our classical solvers reach within fixed budgets.

We study alternateness of the Jones-Kauffman polynomial of virtual links. In particular, we identify a class of virtual links with alternating Jones-Kauffman polynomial that will be called the class of mock-alternating virtual links. Then, we explain how every mock-alternating virtual link can give rise to an infinite family of mock-alternating virtual links using the twisting technique introduced in [4] and later generalized in [5,18]. Finally, we give a connected state expansion of the mock-determinant for checkerboard-colorable links.

Matrix completion underlies problems from tabular imputation to causal inference, yet existing tabular foundation models treat it as entry-by-entry prediction, repeating context for every target and discarding the matrix's two-dimensional structure. We introduce MatrixFormer, a pre-trained matrix-native transformer that predicts a full distribution for every missing entry in a single forward pass. MatrixFormer is trained entirely on synthetic low-rank and latent-factor matrices under diverse missingness patterns. Applied zero-shot and with the same model weights, MatrixFormer achieves competitive performance on causal inference panel-data tasks, language-model benchmark-score completion, tabular imputation, and recommendation systems matrix completion. These results position MatrixFormer as a general-purpose foundation model for matrix completion.

Recurrent-attention hybrid language models (LMs), which interleave attention and recurrent layers, are increasingly used to combine the efficiency of the recurrent layers with the strong performance of attention layers. Prior work suggests that attention and recurrent layers offer complementary pathways to use past information: attention supports precise memory recall from earlier tokens, while recurrent layers support consolidation of disparate information over long contexts. However, we observe that simply having access to both pathways does not mean that hybrid LMs are effectively using them. We find that they rely substantially more on attention than on the recurrent state. Standard supervised fine-tuning improves overall performance but does not improve how the two memory pathways are coordinated: the model becomes more reliant on information propagated by attention layers, while its use of information propagated by recurrent layers remains limited. To encourage better coordination between the two memory pathways, we add an auxiliary loss that limits attention's access to earlier context while the recurrent state propagates through the full sequence. This objective encourages the model to retain and use information through the recurrent pathway alongside attention. It improves overall performance, with particularly strong gains on tasks involving longer contexts or requiring information aggregation, consistent with the strengths of recurrent layers observed in analysis. Crucially, this imbalance and the benefit of our auxiliary loss generalize: they apply to multiple recurrent-attention LMs in question-answering and agentic tasks, as well as to attention-based LMs that combine different forms of memory. Together, our findings show that simply providing multiple memory pathways does not ensure their effective use, and that targeted supervision is needed to better coordinate them.

Quantum low-density parity-check (qLDPC) codes offer a promising route to reducing the qubit overhead of fault tolerance, with recent ultra-high-rate constructions achieving encoding rates above one half. In this work, we develop fast and parallel logical computation methods for ultra-high-rate codes. We first utilize the group symmetries of the codes to partition the logical qubits into orbits, and then apply parallel surgery methods to implement group-invariant collections of logical Pauli-product measurements. These measurements are highly programmable, as we may arbitrarily couple logical orbits as long as each measurement is repeated across the whole orbit(s). To reduce the space and time overheads, we develop a technique called chain-map convolution, which uses the group structure to boost both the spatial and temporal distances of our protocol in a highly efficient manner. As case studies, we construct distance-preserving parallel measurement gadgets for the [[256, 68, 12]], [[576, 148, 18]] pair-partition codes, and a [[1152, 580, 12]] rate-$1/2$ code, with only around $2.5\times$ space overhead and 3-4 rounds of syndrome extraction each. The same methods also enable parallel injection of cultivated magic states into high-rate codes. Finally, we extend programmability through subgroup restrictions, revealing a tradeoff between logical control and the spacetime resources required for fault tolerance and highlighting the crucial role structured parallelism plays in reducing overhead. These results provide new methods for designing qLDPC logical operations with low spacetime overhead, opening a route toward more efficient fault-tolerant quantum architectures.

In decentralized consumer-to-consumer (C2C) marketplaces, people list goods, negotiate with strangers, and rate one another, so trust rests on reputation. Large language model (LLM) agents now act for users, raising risks to their money, privacy, and reputation. We introduce BazaarBench, a simulated C2C marketplace and benchmark for evaluating the safety of these agents. It tracks ownership, item condition, and commitments across transactions, combining record checks with rubric-based LLM judgments to identify six failure types across five stages. We run three base markets for 30 simulated days, each with 100 agents using one model and inventories drawn from a public eBay sample. Across 45 continuations, we evaluate five models under ordinary instructions, deadline pressure, or adversarial instructions to exploit other traders. Each continuation runs for seven simulated days from a copy of a market's day-30 state. The tested model controls the same 20 selected agents, retaining their personas, inventories, and histories, while the other 80 keep the base model. All five models attempt to promise the same item to multiple buyers under ordinary instructions. Adding targets and deadlines increases these attempts for every model. Under adversarial instructions, the share of tested sellers' committed transactions completed despite unavailable items or overstated conditions rises from 15.4% to 33.4%, reaching 55.5% for GPT-5.4. Averaged across models and markets, simulated weekly earnings per tested agent rise from USD 20 under ordinary instructions to USD 33 under adversarial instructions. Most of the increase comes from items the sellers never held. We release the simulator, saved market states, evaluation code, and records covering 357,608 agent model calls for evaluating new models and developing safer marketplace agents.

We completely determine the possible parities of the Euler characteristic of closed $Pin^{\pm}$-manifolds of a given dimension. As an application we partially calculate the controlled cut-and-paste groups $SKK_{2k+1}^{Pin^{\pm}}$, which were previously undetermined.

This paper is primarily concerned with the existence of a symplectic $4$-manifold, homeomorphic but not diffeomorphic to $CP^2$, which admits a nontrivial finite order symplectomorphism. We reduced the existence of such an exotic $CP^2$ to a question in contact geometry. More concretely, we showed that there are precisely $12$ plumbed manifolds, each equipped with a canonical contact structure associated to the plumbing, such that an exotic $CP^2$ with a nontrivial finite order symplectomorphism exists if and only if one of the $12$ contact plumbed manifolds admits a certain specific symplectic filling. As the main technical result, we completely determined the order and local representations of a pseudofree symplectic $Z_p$-action for an odd prime $p$ on a rational homology $CP^2$ with positive canonical class. Our result suggests that potentially, there are symplectic $Z_p$-actions which are not a finite automorphism of a fake projective plane. Moreover, we introduced a construction, which, under the minimal assumptions on the symplectic fillings of the $12$ contact plumbed manifolds, produces a symplectic $4$-manifold on the Bogomolov-Miyaoka-Yau line via a reverse-engineering process.

Hierarchical graphs embed in hyperbolic space with lower distortion than in Euclidean space owing to its negative curvature. However, their gradient-based learning is hampered at large radii, where the Poincaré ball and the Lorentz hyperboloid models fail numerically. Polar coordinates avoid this problem, but the hyperbolic metric scales the angular step by the hyperbolic sine of the radius, freezing angular motion. We observe that this factor is a choice, silently fixed by existing implementations: the Euclidean tangent parametrization, for instance, uses the radius itself. We show that other choices are not only possible but preferable. They are endpoints of a one-parameter family of optimization preconditioners with curvatures from $-1$ to $0$, while the embedding remains at curvature $-1$. We show that since the Euclidean preconditioner rearranges a layout but refines it poorly, while an intermediate one refines far better once a layout is in place, combining them in two stages reduces the loss on real-world trees by 46-74% over the best single curvature.

ufakzeka-karar is an open Turkish decision model with 182,494,466 parameters. Given a Turkish text and questions of a fixed answer type (a choice, a level on an ordered scale, or yes or no), it returns a temperature-scaled probability for every option and an expected error that serves as a "not sure" signal, without generating text and in one CPU forward pass for up to ten options. Built on the lab's ufakzeka-1-base, its head scores each option blind to the others at shared positions, so the answer does not depend on option order. A sequential head trained with shuffled options was about as accurate but changed 2.3 to 2.8 percent of its answers when only the option order changed; REINFORCE lost 10.2 points (0.102) of macro F1 to cross-entropy. On the open set of HakemBench v1.0 (4,275 questions, 7 tracks) the released model ranks 7th of 16 rows with a composite of 0.660 (95% interval 0.642 to 0.677). Temperature scaling lowers calibration error (smooth ECE) on the development set but raises it on held-out support questions, from 0.027 to 0.045 for the first scored run, which never trained on them; the released model later trained on them, so its 0.036 to 0.064 is not an unseen-question test. The released model is the last of three runs scored on HakemBench, and its numbers are not blind. The second run's new training data was aimed at the first run's errors on the full test set in guardrails, moderation and customer support, and the released run was trained after the second run's guardrail results on the full test set were read, under a protocol fixed in writing before any of its data, code or runs. All its numbers come after these readings; its guardrail, moderation and customer support numbers carry the flag "shaped by reading the test results". With every model scored on the other four tracks only, its composite is 0.678, 6th of 16. Weights and code are under Apache-2.0.

We introduce recurrent incidence automata (RIAs), a new automaton model motivated by a decomposition of certain two-stack visibly pushdown computations. The decomposition separates vertex-local finite-state computations from recurrent one-stack interfaces connecting consecutive vertices. The construction is motivated by a two-stack visibly pushdown encoding of arbitrary ordered graphs whose strings admit a unique factorization into center-foldable vertex-local factors and whose auxiliary stack is empty at every factor boundary. Folding each factor into a sequence of pair symbols yields a local interface transformation. An RIA consists of a finite-state unit that computes these transformations and a recurrent layer that composes them across consecutive factors. Rather than manipulating an internal pushdown store, RIAs externalize long-range stack memory into recurrent interfaces between local computations. We show that nondeterministic RIA languages are closed under union, intersection, concatenation, Kleene-*, and reversal. Deterministic RIAs are closed under Boolean operations, although emptiness remains undecidable.

We study a variant of Schumacher compression. The task asks for the minimal memory cost for compressing many copies of a density matrix with known spectrum and unknown eigenbasis, without preserving its purification. For fixed dimension and distinct positive eigenvalues, we obtain the cost through its additive constant. Achievability uses a generalization of Werner's cloning map to $\mathrm{GL}(d,\mathbb C)$ irreducible representations, with finite trace-distance bounds controlled by highest-weight differences and row gaps. The matching converse follows from a quantitative form of Koashi–Imoto incompressibility for irreducible group orbits under a spectral-gap assumption. We also identify the relation between this quantum minimum description length and universal lossless coding overhead, and its connection to free entropy is explained in a companion letter. We provide a Lean certificate for our proofs.

In free probability, Voiculescu's free entropy counts matrix microstates that approximate an operator. It is a non-commutative analog of the Shannon entropy, distinct from any quantum entropy of a density operator, but it has had no physical meaning in quantum theory. We define a finite-dimensional physical free entropy at finite resolution and show that it directly characterizes, up to an explicit constant, the minimal quantum memory needed to program many copies of a state with known eigenvalues and unknown eigenbasis. This task is a variant of Schumacher compression, and we call its optimal cost the quantum minimum description length. We briefly discuss how free entropy differs from quantum entropies and its potential applications in quantum physics.

Since the introduction of the complexity class QMA as a quantum-verifier analogue of NP (Kitaev, 1997), many have wondered whether quantum proofs are necessary or classical proofs suffice - that is, whether QMA = QCMA or QCMA != QMA (Aharonov and Naveh, 2002; Aaronson and Kuperberg, CCC '07). This longstanding question was recently answered by works of Bostanci, Haferkamp, Nirkhe, and Zhandry (STOC '26) and Bostanci, Huang, and Vaikuntanathan (FOCS '26), which showed that quantum proofs are more powerful than classical ones in the classical-oracle setting. However, it remains unclear what exactly makes quantum proofs more powerful than classical ones. In the information-theoretic setting, a family of quantum states is not classicalizable if and only if it is unclonable. Indeed, these recent works also explicitly highlight the unclonability of their quantum proofs as a key mechanism behind their separations, and their arguments crucially rely on this property. This raises the question of whether unclonability is necessary for quantum proofs to be more powerful than classical ones. In this work, we show that even clonable quantum proofs can be more powerful than classical ones relative to a classical oracle by constructing a classical oracle O such that QCMA^O != ClonableQMA^O. This resolves the open question of Nehoran and Zhandry (ITCS '24), who established the analogous separation relative to a quantum oracle. We also show a classical-oracle separation between BQP/clonableqpoly and BQP/poly, and give applications of our results to quantum cryptography.

We study the dimension defect of finitely many subspaces over an arbitrary field. We derive a symmetric, nonrecursive Grassmann-type formula for the dimension of their sum. The formula expresses the total dimension loss through intersections of a distinguished subspace with partial sums of the remaining subspaces, with coefficients determined by the number of subspaces involved. We also show that the formula admits a Shapley-value interpretation for the associated representable polymatroid. Grouping the correction terms by the number of participating subspaces yields a nonnegative dimension-defect profile. We prove that this profile is monotone and identify its successive gaps with the discrete curvatures of the average-rank profile. These gaps give exact remainders in two-sided defect bounds. Equality in either bound holds precisely when the images of the subspaces in the quotient by their common intersection form an internal direct sum. We also give an exact geometric decomposition of the Kinser slack into nonnegative quotient dimensions and characterize equality. Averaging these slacks over permutations and contractions recovers every defect curvature except the final one; additional representability constraints remain in the individual ordered slacks. Weighted and dual formulas accompany the expansion, and entropy analogues express the defect levels and curvatures as averages of mutual and conditional mutual information, respectively.

In the standard three-flavor paradigm, two of the three neutrino mass-squared differences are independent. The existence of an additional dependence between these oscillation parameters was recently predicted on empirical grounds, taking the form $$\frac{\sqrt{\Delta m^2_{31}}+\sqrt{\Delta m^2_{21}}}{\sqrt{\Delta m^2_{31}}-\sqrt{\Delta m^2_{21}}}=\sqrt{2}.$$ The JUNO Collaboration has now reported the most precise measurements to date of both $\Delta m^2_{21}$ and $\Delta m^2_{31}$. We demonstrate that this relation is in excellent agreement with the experimental data, which yield $1.4143_{-0.0038}^{+0.0036}$ corresponding to a relative precision of $0.27\%$. Remarkably, the central value differs by less than $0.03$ standard deviations from the prediction ($\sqrt{2}\approx1.4142$). This provides strong experimental support for the novel constraint and thereby suggests an underlying symmetry governing the neutrino mass spectrum, rather than purely independent parameters.

We consider the computational task of anti-symmetrization of an unknown mixed quantum state: Given iid copies of a rank-$r$ qudit state $\rho=\sum_{i=1}^r \lambda_i |\varphi_i\rangle\!\langle\varphi_i|$, prepare a copy of the state proportional to $\Pi_{\mathrm{anti}}^r (|\varphi_1\rangle\otimes \ldots\otimes|\varphi_r\rangle)$, with $\Pi_{\mathrm{anti}}^r$ the anti-symmetric subspace projector. We first give a simple algorithm for anti-symmetrization that uses $\widetilde{O}(r / \lambda_{\min})$ copies of $\rho$, where $\lambda_{\min}$ is the minimum non-zero eigenvalue of $\rho$. This algorithm works even in the streaming setting with small working memory, and it uses only simple quantum operations like controlled-SWAPs and constantly many layers of single-qubit gates. As an application of this streaming anti-symmetrization procedure, we give a reduction from gapped mixed-state one-way state generators to pure-state one-way state generators. Then, using tools from Schur-Weyl duality, we give and analyze a second (non-streaming) anti-symmetrization procedure that uses $O(r / \lambda_{\min})$ copies of $\rho$, which then serves as a building block in new algorithms for weak Schur sampling and unitary Schur sampling, outperforming existing approaches in some regimes of number of copies $n$ and local dimension $d$, as well as in an algorithm for optimal purity amplification for qudit states.

Partial distinguishability affects multiphoton interference, which is central to large-scale photonic experiments such as Boson Sampling and to applications of multiphoton states across quantum technologies. Characterizing the internal degrees of freedom responsible for this behavior is challenging: in many scenarios they cannot be accessed directly and even their dimension or physical nature may be unknown a priori. Here we introduce Gram-matrix tomography for independently prepared photons with pure or nearly pure internal states. For pure states, the Gram matrix contains all information required to predict the statistics of experiments that are insensitive to the internal degrees of freedom. We reconstruct this matrix, up to physically irrelevant phase choices, from a quadratic number of measurable Bargmann invariants (quantities encoding overlap magnitudes and collective phases). We access these quantities using two complementary approaches: targeted Fourier interferometers and two- and three-body photon-number correlations measured in a single random interferometer. We derive explicit error bounds that control the total-variation distance between predicted and actual outcome distributions for any subsequent experiment that is insensitive to the internal states. These guarantees account for finite sampling and deviations from purity. Crucially, the total sample complexity of the protocol is independent of the internal Hilbert-space dimension and polynomial under mild assumptions on overlaps of internal states. Our protocol uses standard interferometric measurements and turns them into a robust and operationally grounded characterization of partially distinguishable photons.

For any Galois field $\mathbb{F}_q$ with $q$ elements and any positive integers $m$ and $n$, the directed graph $D(q;m,n)$ has vertex set $\mathbb{F}_q\times\mathbb{F}_q$, and there is an arc from vertex $(x_1,x_2)$ to vertex $(y_1,y_2)$ if and only if $x_2+y_2=x_1^my_1^n$. It was conjectured in earlier work that two digraphs $D(q;m_1,n_1)$ and $D(q;m_2,n_2)$ are isomorphic if and only if there exists an integer $k$ relatively prime to $q-1$ such that $m_2\equiv km_1$ and $n_2 \equiv kn_1$, where both congruences are modulo $q-1$. We prove this conjecture over prime fields and construct an infinite family of counterexamples over extension fields.

The measure of indirect CP violation in the neutral kaon system, $\epsilon_K$, is an important precision test of the Standard Model. A first-principles Standard Model prediction of $\epsilon_K$ requires evaluating the long-distance contribution in which the $\Delta S = 2$ process proceeds via two $\Delta S = 1$ operators. The non-perturbative nature of QCD in the relevant energy regimes requires a lattice QCD simulation. In this paper we present the two-loop, $O(\alpha_s)$ RI/SMOM - $\overline{\text{MS}}$ matching factors required to renormalize a lattice QCD calculation of the long-distance contribution to $\epsilon_K$.

A major practical obstacle to Regev's factoring algorithm is its large space usage. In particular, the algorithm was believed to be incompatible with "qubit recycling," which streams the control qubits in phase estimation rather than storing them all simultaneously. We show how to fix this streaming problem. We build on the construction of Ragavan and Vaikuntanathan, which used Fibonacci accumulators to reduce the qubit count from $O(n^{1.5})$ to $11.32n$. We make two key improvements. First, we use superposition masking to prevent unwanted information from entering the accumulators, removing the need for uncomputation. Second, we show how to perform phase estimation using Fibonacci powers of a unitary operation rather than powers of two. This avoids various encoding conversions required by the construction of Ragavan and Vaikuntanathan. After these improvements, the phase-estimation qubits are used only once in the circuit, allowing them to be streamed. Our improvements reduce the space usage of Regev's factoring algorithm to $n+M(n)+O(\sqrt{n})$, where $M(n)$ is the space required for an in-place multiplication. For example, using the in-place multiplier of Luo et al. (2026), Regev's algorithm can factor an $n$-bit integer using $4n+O(\sqrt{n})$ qubits. Although our construction still falls short of the $n/2+\epsilon n$ qubits sufficient for leading implementations of Shor's algorithm, it is the lowest-space instantiation of Regev's algorithm to date and suggests that Regev's algorithm is still far from fully optimized.

We report a comprehensive investigation of the magnetic, electrical, and thermodynamic properties of single crystalline Tb6Ti4Al43, in which Tb atoms form a distorted kagome lattice embedded within a complex Ti-Al framework. Tb6Ti4Al43 exhibits pronounced magnetic anisotropy and antiferromagnetic ordering at approximately 19 K. For magnetic field applied along the c-axis, a strong metamagnetic switching is observed, accompanied by a clear asymmetric hysteresis at low temperatures, while for field applied along the ab plane, a spin-flop transition is observed. The effect of field induced magnetic states is strongly reflected in magnetotransport measurements, giving rise to pronounced anomalies in magnetoresistance and an anomalous Hall contribution. A significant intrinsic contribution to the anomalous Hall conductivity is observed, highlighting the Berry curvature effects arising from the system's momentum-space topology. Our results highlight the interplay of spin, charge, and lattice degrees of freedom in Tb6Ti4Al43 and establish this distorted kagome antiferromagnet as a promising platform for field-tunable multifunctional behaviour, including electrically readable magnetic switching, spin manipulation, and magneto-mechanical functionalities.

We establish staircase phase transitions for the largest eigenvalue of heavy-tailed sample correlation matrices formed from a $p_n \times n$ data matrix with i.i.d. real entries of mean zero and unit variance, allowing an infinite fourth moment. In the proportional regime $p_n / n \to \phi \in (0, \infty)$, the first-order asymptotics depend jointly on the aspect ratio and the entry tail. The transitions are driven by collisions of large entries in distinct rows of a common column. The first collision order capable of producing a separated upper outlier is $k_* (\phi) = \lfloor \sqrt{\phi} \rfloor + 2$, yielding the critical tail exponent $\alpha_* (\phi) = 2 + 2 / k_* (\phi)$. This exponent decreases in steps as $\phi$ increases, creating a staircase boundary between convergence to the upper Marčenko–Pastur edge and successive outlier levels. At exact critical tail scales, the point process of eigenvalues above the upper edge or the preceding deterministic level converges to a Poisson point process. The resulting nondegenerate limiting laws for the largest eigenvalue connect adjacent phases and have a positive atom at this baseline. If every fixed collision order is supercritical, the largest eigenvalue diverges in probability despite finite entry variance.

The Eastin-Knill theorem rules out a universal set of transversal logical gates on any code that detects single-qubit errors. The theorem constrains one partition of the qubits at a time. We show that depth-one layers of few-qubit gates, each respecting its own partition, can together generate a universal logical gate set on an error-correcting code. Each logical gate generator is then a single depth-one layer followed by error correction of the same stabilizer code, with no magic-state distillation, code switching, gauge fixing, lattice surgery, or other gadgets. We show that such codes can be found by applying an entangling circuit to a lower-distance "seed" code admitting a depth-one non-Clifford gate and an appropriate "kernel" code. We construct 47 codes with depth-one generating sets that tolerate at least one faulty gate per layer, including a [[37,1,7]] and a [[59,1,9]] code, both of check weight at most eight.