Papers

Understanding the qubit environment is central to fault-tolerant superconducting quantum computation. Characterization typically relies on relaxation, leaving excitation largely underexplored. Here, we access this information with time-resolved broadband excitation spectroscopy. The resulting qubit excitation spectrum reveals a highly structured landscape, interspersed with cold regions. Combined with postselection, this technique enables full reconstruction of the noise power spectral density (PSD) and separates quantum from classical noise. With feed-forward, it exposes long-lived two-level-systems (TLSs), whose relaxation times span tens of microseconds to milliseconds - uncovering an intrinsic link between the qubit-TLS coupling and TLS relaxation. The data suggest that the long-lived TLSs are intrinsic to the qubit environment, and can cause excitation that lasts for many qubit operation cycles. Consequently, the environment retains a memory of prior dynamics, and properties like gate fidelity become non-Markovian and protocol-dependent. The presented approach enables identifying and bypassing the hidden roadblocks formed by long-lived TLSs in fault-tolerant quantum computation.

We study synchronized boundary and kernel approximation for harmonic measure systems satisfying nested mean-value identities. Under boundary concentration along a common subsequence, joint limits yield representation on a common refinement, and gluing gives uniqueness on the minimal boundary. Neighborhood nonvanishing gives full-sequence synchronized selection. For a compact set whose removal from a bounded $C^{1,1}$ domain in dimension at least three leaves a connected domain, zero Newtonian capacity is equivalent to concentration of one normalized singular kernel, constructed from harmonic measures, on a subsequence of one relatively compact exhaustion. For such zero-capacity sets, capacitary estimates give quantitative recovery of finite signed source measures on every prescribed relatively compact exhaustion. On rectifiable curve networks in three dimensions, the logarithmic error bound is sharp for this recovery formula in the class of finite measures. Boundary splitting and finite weighted metric graphs illustrate the gluing conclusions.

We give an explicit quasi-isometry from the flip graph of triangulations $\mathscr{F}_{g,n}$ to the $\epsilon$-thick part of Teichmüller space equipped with the Teichmüller metric, by mapping each ideal triangulation to the hyperbolic surface whose shearing coordinates along that triangulation are all equal to zero. Extending this construction, we obtain a quasi-isometry $Q$ from the hexagon graph $\mathscr{H}_{g,n}$ to the augmented Teichmüller space $\overline{\operatorname{Teich}}_{g,n}$ equipped with the Weil-Petersson metric, and we estimate its width, that is, the Hausdorff distance between $Q(\mathscr{H}_{g,n})$ and $\overline{\operatorname{Teich}}_{g,n}$. By bounding the Weil-Petersson distance along grafting rays, and by using an analogue for ideal triangulations of Wolpert's upper bound on the Weil-Petersson distance to a stratum, we show that the width is at most of order $\sqrt{(g+n)\log(g+n)}$. We also provide an explicit projection from any point in $\overline{\operatorname{Teich}}_{g,n}$ to the thick part of boundary strata, maintaining simultaneous control over the Weil-Petersson distance and the shearing coordinates.

P-type transparent conducting materials (TCMs) remain scarce, limiting progress in transparent electronics and tandem photovoltaics. Here, we perform a high-throughput computational search across CuI-based ternary X–Cu–I compounds using the minima hopping method combined with density functional theory. Filtering for thermodynamic stability, optical transparency, and light hole effective mass, we identify 58 candidate p-type TCMs, of which 49 are previously unreported. Twenty are p-type degenerate semiconductors with intrinsic hole carriers, including Cu$_2$FI$_2$ (calculated gap 3.75 eV, hole effective mass $0.262\,m_0$). We further use this dataset to revisit the chemical modulation of the valence band (CMVB) design rules, originally formulated for Cu-based delafossite oxides. Four generalizations emerge: linear Cu coordination is not required, as tetrahedral coordination (CN = 4) is compatible with p-type TCM character; the mechanism extends to halide-based systems with I $5p$ anion states; the Cu $d$ state energy center $\epsilon_d$ is a quantitative descriptor for valence band dispersion; and Cu undercoordination in monovalent-anion hosts drives intrinsic p-type carrier generation through stoichiometric charge balance. These results establish a generalized computational framework for the design of p-type halide TCMs.

We present a Banach space $X_\alpha$, equipped with an explicit norm, whose properties resemble those of the predual $JT_*$ of the James Tree space. In particular, $X_\alpha$ admits a shrinking dyadic basis with the property that the sums of its elements over any segment are uniformly bounded. The space $X_\alpha$ is complementably $\ell_2$-saturated, and its dual $X_\alpha^*$ does not contain an isomorphic copy of $\ell_1$. Finally, the bidual $X_\alpha^{**}$ is also complementably $\ell_2$-saturated, and the quotient $X_\alpha^{**}/X_\alpha$ is isometric to $c_0(2^{\mathbb{N}})$.

Expert-driven Bayesian networks can support recurring software decisions when historical data are limited, but quantifying their conditional probability tables (CPTs) requires many probability judgments. Semi-automatic methods reduce direct elicitation, yet practitioners have little comparative evidence from operational decision processes. We report an embedded case study comparing the Weighted Sum Algorithm (WSA) and Ranked Nodes Method (RNM) as complete elicitation-and-quantification pipelines in two contexts within a single software research and development (R&D) organization: feature selection for Internet of Things projects and user interface design selection. Within each context, the pipelines shared the graph, root priors, and decision evidence. We evaluated them through 15 expert-defined model-walkthrough scenarios and retrospective reconstructions of alternatives recorded in decision meetings. WSA matched 7/7 and 6/8 walkthrough expectations, whereas RNM matched 4/7 and 4/8. Both pipelines placed the selected alternatives near the top of the retrospective rankings. Their complete distributions nevertheless differed: the median total variation distance was 0.304 across four distinct feature patterns and 0.340 across seven distinct design patterns, rising to 0.890 for one design pattern. These results concern two ordinal value-estimation models in one organization. They show that similar shortlist behavior does not imply equivalent representations of uncertainty. For comparable software decision-support models, method selection and validation should consider node semantics, the judgments experts can provide, calibration requirements, complete output distributions, and the intended downstream use of the probabilities.

We develop foundations for metastable homotopy theory using Heuts' 2-truncated Tate coalgebras.

We revisit the correspondence between three-dimensional gauged linear sigma models and K-theoretic Gromov-Witten theory. We show that the K-theoretic Hori-Vafa formula for the grassmannian arises from the three-dimensional gauge theory upon including an appropriate anomaly-cancellation factor. Furthermore, we derive K-theoretic I-functions for isotropic grassmannians by incorporating symmetric and anti-symmetric tensor representation fields together with suitable boundary conditions.

AI is becoming increasingly integrated into daily workflows, especially in computing. We are gradually shifting towards an AI-rich future, an impending yet unknown one. One important emerging concern is whether we are accordingly preparing our future computing workforce. Further, we need to know what the important cognitive skills are to remain relevant in the computing workforce and if there are changes in cognitive skill importance. To investigate this direction, we conducted a mixed-methods study, collecting perceptions from computing students and computing experts regarding the importance of cognitive skills in the past, present, and future. We report that the perceived importance of most cognitive skills will decrease in the future, with an AI-rich environment, but critical thinking skills remain important. Further, we report reasons collected through interviews on why the importance of cognitive skills will change and how future computing students can prepare for it.

Particle-resolved aerosol models preserve the joint size-composition structure that governs cloud activation, optical properties, and freezing, but their computational cost limits their use in large-scale atmospheric models. We extend AeroMELD from a compact representation of aerosol populations to a prognostic model of coagulation. The resulting AeroMELD-Coag advances nine learned coordinates for population shape and one for total particle number through a scale-covariant neural ordinary differential equation that incorporates the known concentration dependence of coagulation. Across 2,000 held-out trajectories spanning 48 h, median symmetric errors are 3.5% for latent shape and 0.52% for total number. This ten-coordinate state also retains the evolving size-resolved composition and associated cloud-condensation-nuclei (CCN) activation, optical properties, and frozen fraction. Compared with the evaluated 20-bin sectional model with 320 prognostic coordinates, AeroMELD-Coag lowers pooled mean CCN activation error from 1.98% to 1.38% and frozen-fraction error from 1.72% to 0.56%. In a matched integration benchmark, GPU integration time per trajectory, amortized over a batch, is more than three orders of magnitude lower than for both single-core CPU references. These coagulation results provide a proof of concept for advancing aerosol microphysics efficiently in a compact learned state while retaining information about aerosol mixing state. They establish a foundation for future large-scale atmospheric models to represent aerosol microphysics with greater mixing-state detail within practical computational budgets.

Polymers are widely used as functional insulating materials, but predicting their dielectric behavior is challenging because multiple mechanisms, acting across different spatial and temporal scales, contribute to their response. In polyethylene (PE), one of the most common polymers, electronic, atomic, and mesoscale dynamics must all be considered. In this work, semicrystalline models of polyethylene were developed and investigated using a multiscale approach. Quantum simulations were employed to describe electronic and vibrational contributions to the dielectric response, while classical molecular dynamics simulations captured the behavior of polymer chains at room temperature and frequencies down to the 10-100 MHz range. The study focuses on the impact of radio-oxidation defects on the dielectric properties of polyethylene. Quantum calculations reveal how electronic and vibrational contributions depend on the local atomic environment surrounding these defects. Comparison with molecular dynamics results highlights the influence of temperature on the static dielectric constant. Structural analyses further assess the effect of each defect type on PE crystallinity. Analysis of dipolar correlation functions shows that different defects interact with one another and with the polymer matrix in distinct ways, affecting permittivity and dielectric loss peaks. For all defect types, defect-defect interactions provide the largest contribution to dielectric response. While most oxidized groups exhibit positive coupling with the PE matrix, alcohol defects display a negative cross-correlation, partially offsetting their impact on the dielectric properties. Our results also show that ketone groups produce the largest dielectric loss between the defects studied.

We study the Goodwillie tower of the identity evaluated at the circle. Our main result is the p-local equivalence $\Omega P_{p^k}I(S^1) \simeq \mathbb{Z} \times \Omega^{\infty+2k}\tau_{>0}\mathbb{Sp}^{p^k}$ obtained from the work of Behrens and Kuhn on the Whitehead conjecture. In particular, this identifies the Goodwillie approximations of the circle as infinite loop spaces, a phenomenon which does not hold in general. The equivalence also reduces the computation of their homotopy groups to a stable problem, making them accessible to tools such as the Adams spectral sequence and to computer calculations. We place these computations in a more general framework by developing a Goodwillie-calculus analogue of the Gray sequence, which relates the homotopy groups of successive stages of the tower.

We show that a linear damped elastic system can display a Mpemba-like reversal of vibrational relaxation without any nonlinear constitutive response or amplitude-dependent damping. The mechanism is purely modal. For a one-dimensional Kelvin–Voigt elastic medium, the decay rate of the $n$-th vibrational mode scales as $\gamma_n\propto n^2$. We formulate relaxation in terms of a positive modal envelope energy, thereby excluding crossings caused merely by oscillation phase. A one-parameter family of initial states is then constructed such that increasing the initial elastic excitation simultaneously reduces the projection onto the slow fundamental mode. In the invariant two-mode subspace, the relaxation-order crossing is obtained analytically and, for this family, occurs at a universal dimensionless time independent of the selected pair of initial states. In the limiting case where the slowest mode is absent, the dominant relaxation rate changes discontinuously, yielding a direct vibrational analogue of the strong Mpemba effect. We further show that the same relaxation-order reversal occurs in the ordinary mechanical vibrational energy, which decreases monotonically despite retaining the oscillatory dynamics of the underlying modes.

Recent advances in experiments using individually trapped atoms have enabled precise control over circular Rydberg electrons with exceptionally long lifetimes. We show that these giant and stable electron orbits can form toroidal lattice potentials for ultracold atoms with a period set by the electron's de Broglie wavelength. Unlike conventional static optical lattices, this electron lattice is formed via the electron-atom interaction, which mixes Rydberg circular states with opposite azimuthal phase winding into a standing electronic matter wave. The lattice phase is thereby intrinsically coupled to the atom position. For a pair of atoms, this results in ballistic tunneling motion in the two-atom spatial correlations along the ring, while the single particle dynamics is essentially free rotation. We simulate the dynamics for an experimentally realistic setting that exploits optical tweezers for individual atom control. Our results open a route toward incorporating long range atom-atom interactions mediated by a single electron and, ultimately, realizing small Bose and Fermi gases confined in these microscopic electronic atom traps.

Ground-state preparation is a fundamental challenge in quantum many-body physics, chemistry and materials science. Dissipative state-preparation techniques engineer a Lindbladian coupling to a bath that drives the system to its ground state, needing no initial overlap. Their cost is dominated by the Hamiltonian simulations inside the filtered jump operators, since each jump is a weighted sum of Heisenberg-evolved copies, or branches, of a coupling operator. We reduce this cost by following three simple observations to their logical end. First, on a fixed time grid, linear programming optimises the branch weights, trading the heating response against the circuit cost. Second, Lieb–Robinson bounds let each branch evolve on a patch whose radius grows with its duration. Third, if the dynamics converges quickly and truncation changes the jumps only slightly, the stationary state stays near the ground state even on patches smaller than the light cone. We validate our approach numerically. On a grid of $81$ evolution times, the optimised filter reduces the worst-case heating response, or leakage, more than $116$-fold at equal cost. Across fifteen molecules and six chains, the cost falls by 27 to 56% at matched design-grid leakage; at matched cost, the molecular leakage falls by a factor of $2.9$ to $2\times10^{3}$. Density-matrix simulations show high ground-state fidelity for both filters, and on Ising and Hubbard chains the cheaper filter keeps a ground-state population of at least $0.89$ on patches of three or four sites. The same patches fail on a Heisenberg chain whose truncation error falls more slowly with patch size and whose Lindbladian gap is 12 times smaller than the Ising chain's, pointing to a role for the gap in how far patches can be truncated. At fixed radius the per-branch cost becomes independent of system size, reducing the cost of dissipative state preparation.

Strong light-matter coupling offers exciting opportunities for controlling molecular properties, yet resolving single-molecule behavior from collective effects remains a challenge. Recent experiments in scanning tunneling microscope break junctions (STM-BJs) have demonstrated strong light-matter coupling at the single-molecule level. Because STM-BJs provide direct access to molecular-junction currents, we investigate whether strong coupling leaves a measurable fingerprint in transport, potentially probing polaritonic states beyond optical spectroscopy. Using a semiclassical mapping approach for nonequilibrium quantum transport, we find that steady-state current indeed carries a direct signature of single-molecule strong coupling. We test the robustness of this signature against electrode connectivity, nuclear motion, spatially structured electromagnetic modes, and solvent effects. While environmental factors and nuclear motion reshape the signature, the current is not a passive response but is actively influenced by coupled molecular dynamics. Our work establishes current as a tool to interpret and control strong light-matter interactions in nanoscale junctions.

With the growing importance of science-related discourse on the Web and the erosion of the classical knowledge order, it is important to identify different user groups, such as scientists, automatically. This work proposes an approach for identifying scientists and non- scientists on X/Twitter based on their user biographies and tweets. We show that we are able to classify accounts as scientists and non- scientists on two different datasets, reaching an F1 score of up to 0.88 using Random Forests with linguistic features and up to 0.96 using a contrastively fine-tuned DeBERTa model in an ensemble setup. Furthermore, we provide two datasets with X users labeled as scientists or non scientists and their respective tweets and user biographies.

Option-implied factors are largely, but not entirely, spanned by the equity factor zoo. We construct 137 option-implied characteristics for optionable U.S. stocks from 2004 to 2023, screen them to 20 representative long-short factors, and test each against 160 equity factors with the double-selection LASSO of Feng, Giglio, and Xiu (2020). In-sample, the equity factors account for about 91% of the option factors' variance, and adding the option factors does not significantly raise the maximum Sharpe ratio. Individually, however, 8 of the 20 option factors, mainly measures of jump tails, kurtosis, and implied-volatility convexity, have non-zero SDF loadings after controlling for the equity zoo, and 4 remain significant after a Bonferroni correction. Options thus add a few pricing-relevant dimensions rather than a new factor zoo.

Cross-client duplicate data in large language model training corpora degrades the efficiency of federated learning (FL) while exacerbating model memorization and privacy risks. Privacy-preserving cross-client deduplication effectively mitigates this issue by eliminating duplicate training data. However, existing schemes all follow a "Deduplication-before-Training" paradigm. This serially coupled paradigm incurs high fault-tolerance costs and lacks support for dynamic client joining. To this end, we propose an unexplored paradigm called "Deduplication-while-Training (DwT)", which enables concurrent deduplication and training. DwT transforms cross-client deduplication from a one-time, globally synchronous preprocessing operation into a continuous online service with state management, concurrent claiming, and failure recovery. By enabling state synchronization and task takeover, it minimizes the impact of client disconnections on the overall training progress while supporting the dynamic joining of clients. We design DwT-FL, a privacy-preserving deduplication system, to support DwT. By designing a concurrent state-claim mechanism and a hot-cold dual-queue scheduling strategy, DwT-FL enables the parallel execution of secure deduplication and model training, while effectively handling client disconnections and dynamic joins. Experimental evaluations demonstrate that, compared to the state-of-the-art scheme, DwT-FL significantly reduces the time overhead of failure recovery and dynamic joining by up to 93.04% and 94.18%, respectively. This provides an efficient and elastic concurrent deduplication scheme for dynamic and unstable FL environments.

The quadratic complexity of dense self-attention remains a central bottleneck for long-context language modeling. Many efficient alternatives address this cost by deciding in advance where attention should be sparse or local. We argue that attention approximation should instead be approached as a geometric problem, with the relevant interaction geometry learned from data: natural-language dependencies are input-dependent and difficult to prescribe in advance, so the model should learn where positional relevance can decay and where broader interactions must be preserved. We introduce Mixture of Semantic Attention Regimes (MoSAR), which learns such an adaptive, controlled-decay geometry over query–key interactions. Input-conditioned query and key routers, applied after positional encoding, select mixtures over short, medium, and global regimes, inducing a continuous distance-dependent attention field rather than a fixed sparsity pattern. This geometry is learned during training and can subsequently be discretized through top-1 routing. In controlled pre-training experiments with matched 500M-parameter models, MoSAR learns a substantially lower-reach attention geometry without degrading language-modeling quality, improving perplexity over dense RoPE at the training context length. Under length extrapolation, MoSAR achieves the best perplexity among all evaluated variants, including strong baselines such as ALiBi. Moreover, the learned geometry remains stable under deterministic top-1 discretization, suggesting that it is not only adaptive, but also amenable to low-cost approximation at inference time.

We investigate the systemic macroscopic dynamics emerging from Limit Order Books (LOBs) populated exclusively by autonomous reinforcement-learning agentic traders. By formalizing agent interactions within a microscopic order-matching engine, we examine two fundamental quantitative phenomena: equilibrium phase transitions in order flow regime shifts, and the structural dynamics of market impact. We show that agentic LOBs exhibit distinct phase boundaries separating orderly price discovery from hyper-volatile cascade states, governed by critical thresholds in the number of agents and observable market depth. Furthermore, we demonstrate that market impact under agentic liquidity provision deviates from classical square-root dynamics, exhibiting distinct dissipative, balanced, and non-dissipative regimes under non-linear feedback loops.

Digital twins are increasingly used in diabetes research, but reproducing individual glucose dynamics requires accurate identification of glucoregulatory model parameters. Traditional sensitivity analysis can identify influential parameters, yet a ranking based on limited conditions may miss parameters that matter during specific disturbances or for particular individuals. We therefore examine both the magnitude and timing of parameter influence across dynamic input-output conditions and assess whether a common ranking holds across participants. We analyze the Hovorka glucoregulatory model using data from 192 participants receiving automated insulin delivery therapy in the Type 1 Diabetes and Exercise Initiative dataset. We extend Sobol sensitivity analysis to time series and rank parameter influence under four conditions: full-day profiles, isolated meal disturbances, insulin bolus injections, and postprandial responses. We combine the condition-specific results into a global ranking and use it to select parameters for participant-specific identification. Compared with population parameters, identification restricted to the sensitivity-derived subset reduces the root mean square error of 60-minute glucose predictions by 60%, to approximately 31 mg/dL. These findings suggest that a global ranking can capture parameter influence across individuals and dynamic conditions. By narrowing the parameters requiring identification, this approach reduces computational cost and could accelerate the development of personalized diabetes digital twins.

Optimizing atmospheric pressure interfaces requires balancing efficient ion transport with unwanted collision activation. The complex environment at the inlet region, including rapid pressure drops, RF/DC electric fields, ion-gas collisions, and space charge effects, makes simulation of this system challenging. In this work, we introduce MuCITE (Multiphysics simulation for Continuous Ion Transport and Energetics), which combines continuous ion injection, precomputed electric and gas flow fields, and particle-in-cell space charge updates. Furthermore, by employing improved impulsive collision theory, it can not only predict transport but also analyze the loss position of ions and the evolution of their kinetic energy, internal energy, and dissociation probability during transport. Transport predictions with ion currents and activation trends of in-source fragmentation were compared with experimental measurements using a custom-built source and a Q-Exactive HF source. The results show that the simulation predictions agree well with the experimental results and can serve as a reference before experiments. MuCITE allows users to derive ion transport, output beam characteristics, and ion energy through the configurable interface geometry and operating conditions. It supports simulation-guided design and tunable parameters to achieve predicted efficient ion transport or in source activation.

Reliable high-dimensional variable selection requires scalable error-controlling methods. The Terminating-Random Experiments (T-Rex) selector estimates the false discovery rate (FDR) by aggregating early-terminated forward-selection paths in which predictors compete with synthetic dummies. We address two remaining challenges: i) predictor dependence can bias dummy-predictor competition; ii) computation is wasted on recomputing terms shared across experiments. Building on memory-efficient virtual dummies, which sequentially sample projections from their exact conditional law, we estimate the conditional Gaussian law of inactive predictors and map draws onto the remaining sphere radius. A shared lazy Gram cache computes the response product and each requested Gram column once across experiments. Simulations across three covariance structures show that uniform spherical dummies can exceed the target FDR, whereas the proposed method empirically controls FDR. Caching yields more than a sixfold speedup, and the method remains feasible with 100000 predictors, where competing FDR-controlling methods become computationally impractical.

Inspired by the planted clique problem for random graphs, we introduce the planted totally-isotropic space problem for random tensors as follows. Let $U\cong \mathbb{F}_q^n$ and $W\cong \mathbb{F}_q^m$ be finite-dimensional vector spaces over a finite field $\mathbb{F}_q$. Given $d\in \mathbb{N}$, choose a random \(d\)-dimensional subspace \(V\leq U\), and construct a random alternating bilinear map $\phi:U\times U\to W$ subject to the constraint \(\phi(V,V)=0\). Such a $V$ is known as a totally-isotropic space of $\phi$, and the goal is to recover $V$. Building on the recent probabilistic analysis of random tensors (Pham–Qiao–Wigderson–Wigderson, in progress), we initiate the study of the algorithmic hardness of this problem. Setting $m=\lceil n/\log n\rceil$, we show that this problem admits an average-case polynomial-time algorithm for $d\geq n/2$, by leveraging recent advances on the non-commutative rank problem. We also show that this problem admits a $q^{O(n\log n)}$-time algorithm. We carry out algorithmic experiments using polynomial-system solving. From these results, we conjecture that the planted totally-isotropic space problem for $d=\lceil n/C\rceil$ with some constant $C\geq 3$ is exponentially hard. Based on this evidence of computational hardness, we explore cryptographic applications of the planted totally-isotropic space problem and related planted tensor problems. We present private simultaneous messages and secret sharing protocols based on planted tensor problems, following the protocols based on planted subgraphs in (Abram–Beimel–Ishai–Kushilevitz–Narayanan, TCC'23). At the same security level, the public information size of protocols based on planted subgraphs is (moderately) exponential in that of protocols based on planted tensors, while the communication costs of these protocols are polynomially related.

General-purpose agent memory summarizes conversations: it extracts salient snippets, embeds them, and retrieves the top-k into the prompt. A health agent cannot run on summaries: a dose becomes a sentence, "since last week" is resolved at the model's discretion, and a three-month glucose trend cannot be answered by text similarity. We present PIA, a personal intelligence agent deployed alongside a consumer health agent. PIA receives the agent's natural-language requests, decides for itself whether and how to write or read, and turns conversations into typed clinical records and records into a synthesized understanding of the user. Its memory harness consists of four controls – extraction, memory, retrieval, and understanding – each a domain-agnostic mechanism with a pluggable health module: schema, medical alias dictionary, knowledge graph, and temporal rules. We show how the same query receives a different answer as the memory injected into the response context deepens from one-dimensional recall, to a two-dimensional health snapshot, to a three-dimensional trajectory with causality, and report lessons from operation: self-reported health data are missing not at random, question phrasing governs the quality of synthesized understanding, and nearly a third of candidate causal links are structural noise that rules alone remove.

We present a general framework for simulating the nonequilibrium vibronic dynamics of molecules interacting with metal surfaces, which utilizes non-Markovian electronic friction and the corresponding generalized Langevin equation. The method employs a Markovian embedding scheme to sample coordinate-dependent quantum colored noise, with the underlying memory kernels represented systematically by exponential decompositions obtained directly from the electronic forces. In doing so, the approach extends the widely used electronic friction and Langevin dynamics formalism beyond the conventional Markovian approximation while retaining the coordinate dependence of the electronic back-action. We demonstrate the method for nonequilibrium charge transport through a vibrationally coupled molecular junction, where comparisons with numerically exact hierarchical equations of motion simulations show that the non-Markovian formulation is both more accurate and more robust than its Markovian counterpart. In particular, we find that non-Markovian electronic forces play a decisive role in the nonequilibrium vibrational dynamics, stabilizing regimes in which the conventional Markovian approximation fails.

Sequential recommenders typically rely on learnable positional embeddings to encode the order of user interactions. In this work, we ask whether this ordinal signal can be replaced by a structural one derived from the item space. We propose to use Laplacian positional embeddings in SASRec: we build an item co-occurrence graph from training interactions, compute eigenvectors of its symmetric normalized Laplacian, and use them as frozen graph-derived positional embeddings. The backbone architecture and training objective remain unchanged. Experiments on four public sequential-recommendation benchmarks show that this simple replacement improves SASRec performance on most ranking metrics and remains competitive with strong positional and temporal encoding baselines. These findings indicate that item-item graph structure can be an effective substitute for standard ordinal positional embeddings in sequential recommendation.

The arrival of quantum computers threatens the security guarantees of classical cryptography, since quantum algorithms can break schemes that remain secure against conventional attacks. The National Institute of Standards and Technology (NIST) has therefore standardized a set of post-quantum cryptographic algorithms, among them Falcon, a lattice-based digital signature scheme with the most compact signature and public-key sizes of the standardized candidates. Falcon's reliance on floating-point arithmetic makes it hard to implement in hardware, and prior work offers only partial accelerators for specific operations such as signature generation or verification, or a single full implementation generated through high-level synthesis (HLS). This work presents Peregrino, the first hardware accelerator of the complete Falcon digital signature scheme designed from scratch in HDL, targeting resource-constrained edge devices without a native floating-point unit through an emulated floating-point datapath. Implemented on a single Artix 7 XC7A200T FPGA, Peregrino uses 85261 LUTs, 41382 FFs, 44 BRAMs, and 142 DSPs for the Falcon-1024 variant. Against the only prior full implementation, the HLS-based FalconTakesOff, it uses 1,9$\times$ fewer LUTs, 3,4$\times$ fewer FFs, 2,7$\times$ fewer BRAMs, and 9,9$\times$ fewer DSPs, fitting the entire scheme on one FPGA where the HLS design requires at least two. Operating as a peripheral of an on-chip MicroBlaze soft-core, the accelerator additionally reduces key-pair generation, signature generation, and signature verification clock cycles by 92%, 96%, and 85% over the emulated floating-point reference software.

Frequency continuation does not improve every neural representation, and better velocity recovery need not imply better prediction of unseen waveforms. We compare a velocity grid with total-variation regularization, an adapted sinusoidal coordinate network (SIREN), and a frozen image generator with a full-grid residual in two-dimensional acoustic full-waveform inversion. All methods use the same frozen SWEEP discretization, acquisition, observations, background, and budget of 1,500 shot–model gradient evaluations. Three exposed synthetic targets represent inclined layering, curved layering, and a high-contrast curved interface drawn from a fault-labelled family. With 2-to-10 Hz continuation, the generative-residual model attains velocity RMSEs of 218.7, 108.0, and 341.9 m/s, compared with 232.5, 100.3, and 412.9 m/s for Grid+TV. Thus the latter wins on curved layering, while the former improves model RMSE by 5.9% and 17.2% on the other targets. Grid+TV nevertheless predicts held-out 10 Hz shots better on all three. SIREN is worse than the common background in every cell; continuation worsens two targets but improves one. A five-weight TV sensitivity study preserves the CurveFault model-error ranking. An equal-budget variational extension fails both recovery and interval coverage. These results establish conditional rankings, not universal superiority or an isolated spectral-bias mechanism. Full-grid residuals also prevent attributing gains uniquely to the generator. We advocate joint reporting of model recovery, unseen-shot prediction, optimization cost, and negative outcomes when auditing neural FWI.