Papers

The cross-correlation between weak lensing (WL) and the thermal Sunyaev-Zel'dovich (tSZ) effect probes the connection between the matter distribution and the thermal state of baryons at low redshift ($z\lesssim1$). A key systematic in such measurements is contamination by extragalactic foregrounds, particularly the cosmic infrared background (CIB). Here we describe the CIB spectral energy distribution (SED) with a small number of physically motivated spectral parameters, calibrated against external measurements of the radiation field intensity in galaxies. Their uncertainty is then propagated into the recovered WL$\times$tSZ signal in component separation at the level of the multi-frequency cross-power spectra rather than the map level. Applying this method to cosmic shear from the Dark Energy Survey and frequency maps from Planck, we show that the resulting measurements of the WL$\times$tSZ power spectrum are insensitive to variations in the effective CIB spectral index, radio source contamination, and the scale and redshift dependence of the SED. The measurements are robust to realistic foreground uncertainties, supporting their use as probes of baryonic feedback and cosmology, and providing a framework for future high-precision surveys. We compare our measurements with the Flamingo suite of hydrodynamical simulations and find that they are compatible with the Flamingo predictions for a low amplitude of matter fluctuations, parametrised by $S_8$, strongly disfavouring the larger $S_8$ preferred by Planck. This result is independent of the impact of baryonic feedback as modelled in Flamingo.

We study whether a manually specified runtime supervisor can correct recurring failures of an existing end-to-end parking policy in a fixed CARLA parking lot. The vision-based Transformer architecture is inherited from Yang et al.; our contribution is a timed, rule-based Parametric Safety Shield (PSS) applied to its control outputs. The PSS uses hand-calibrated speed, position, and duration thresholds to intervene in observed failure modes, including boundary exits, delayed braking, and stalled or oscillatory control. In the reported closed-loop evaluation, 16 held-out target slots and six initial poses are each evaluated in four rounds (384 attempts per configuration). Target success increases from 327/384 (85.16%) for the retrained policy to 375/384 (97.66%) with the PSS; mean position and orientation errors among successful attempts are 0.21m and 0.33 degrees. These results show an improvement within this simulator setup. The repeated attempts share one map, vehicle, and sensor configuration, and the PSS uses simulator world coordinates; thus the results do not establish generalization to other lots or real vehicles, or a formal safety guarantee.

Changes in listener acoustics require active noise control (ANC) filters to be redesigned for new acoustic paths. We introduce PRIME-ANC, a shared neural synthesizer that learns a bounded, path-dependent logmagnitude correction to a regularized path-ratio base. Minimum-phase reconstruction and truncation produce finite-impulse-response (FIR) filters; training optimizes their noise-control performance. Across ten random training/test splits per dataset, PRIME-ANC achieves average held-out one-third-octave reductions of 18.81 and 17.76 dB over 50 Hz-5 kHz on a ten-path dataset and a public earphone database, respectively. On original earphone measurements, it improves upon the path-ratio base by 7.89 dB. Ablation studies support the contributions of both the analytic base and the path-dependent correction. Given calibrated paths for a held-out listener condition, PRIME-ANC generates a path-specific FIR without iterative optimization. With three Gauss-Newton updates, it reaches 21.43 dB reduction, approaching direct weighted least-squares design while producing lower amplification and root-mean-square control output.

Many planning and control tasks for tendon-driven continuum robots (TDCRs) require the complete Cartesian backbone geometry. We present a reduced Cartesian framework for planar, axially compressible TDCRs that propagates equilibrium configurations along prescribed tendon-force and tendon-displacement trajectories. The backbone is represented by two global position fields. Following exact variation, a Taylor-Galerkin reduction condenses prescribed spatial properties and distributed loads into offline moment vectors, yielding analytic reduced residuals and Jacobians without online spatial quadrature or numerical differentiation. Analytical differentiation and residual correction yield first-order rate systems requiring one fixed-dimensional linear solve per rate evaluation after initial equilibrium alignment on a regular branch. Across four simulated cases covering variable tendon routing, nonuniform geometry, axial compression, and their combined effects, the propagated Cartesian shapes and distributed strains closely match pointwise geometrically variable-strain (GVS) equilibrium solutions. Residual correction suppresses propagation drift across the tested step sizes while adding only about 0.98% to the mean update time of uncorrected Euler. The proposed method requires 0.508 ms per update on average, approximately 11 times faster than pointwise GVS solves. Displacement-driven experiments yield a maximum normalized mean backbone position error of 1.02% and a maximum end-effector position error of 0.28%. These results support efficient and accurate Cartesian full-shape prediction along prescribed actuation paths.

General-purpose multimodal agents can write robot-control programs, but repeated exploration and model-mediated action selection can make execution slow. We study how external body knowledge, successful experience, and executable skills improve an XLeRobot controlled by GPT-6-Astra in a simulated and a physical elevator-button task. In 30 fixed-start simulation trials, complete robot geometry and camera information reduce mean completion time by 57.4% relative to a baseline with only the common control interface and no prior experience; images with synchronized action and state records reduce it by 68.6% without additional body assets. In nine paired comparisons (18 trials) at starts displaced by 10-100 cm, experience recorded at the original start reduces mean time by 58-63% relative to no experience, demonstrating generalization to the tested new starting positions. During experience experiments, GPT-6-Astra spontaneously generates a short visual-feedback program. Researcher-refactored versions reduce mean local-task time by 29-31% in 27 simulation trials. Finally, 12 real-robot trials using operator-confirmed button contact demonstrate sim2real reuse: at a shared nominal start, simulation XML assets and simulation experience reduce mean time by 53.0% and 49.9%, respectively; real experience also transfers to two new starts. These results suggest a practical way to build general-purpose manipulation experiments around GPT-6-Astra: supply machine-readable body descriptions and synchronized demonstrations, and turn useful agent-generated feedback routines into reusable skills, while the agent adapts actions from current images. We release all task prompts, trial-level experimental data, and acquired skill implementations at https://github.com/hesd10/astra-robot-sim2real.

We study the impact of conditional complier average causal effect (CCACE) estimation methods on the performance of subgroup discovery and heterogeneous causal effect estimation under imperfect compliance. Building on the Bayesian Causal Forest with Instrumental Variable (BCF-IV) (Bargagli-Stoffi et al. (2022)) method, we introduce a two-step, model-agnostic approach that allows any suitable machine learning method to be used for the CCACE estimation in the first step. Specifically, we implement two non-Bayesian tree-based methods, both using forest-based learners: DRRF-IV, a debiased transformed-outcome regression-forest approach, and a GRF-based IV adaptation of the generalized random forest framework (Athey et al., 2019). Through a simulation study, we assess the precision, bias, and the ability to correctly identify the underlying subgroup structure of the proposed methods relative to BCF-IV. The results show that the non-Bayesian methods perform competitively across the considered simulation settings, with performance improving for larger sample sizes and moderately large treatment effects, while reducing computational runtime. We apply our new methods by revisiting an empirical study that examines the effect of prompt admission to intensive care units (ICU) on 28-day mortality across 48 UK National Health Service hospitals. While previous work finds no significant overall treatment effect, we investigate whether subgroups of patients may benefit more from prompt ICU admission.

Networking defenses against traffic analysis fill idle slots with cover traffic, yet the physical layer carrying it still hands upper layers the erasure-only model of a silent-idle link, leaving the cover content unexploited. We shape that content: the idle-slot waveform is drawn from the codebook complement under a design distribution, the litter distribution, chosen to defeat a public-design passive observer while the upper layer keeps its block z-channel model. The maximum a posteriori (MAP) receiver reduces to a single threshold on the log-likelihood ratio between the best codeword and the prior-weighted litter aggregate; its errors are silent corruptions the upper layer cannot detect and erasures that retransmission absorbs. Receiver design is thus one-dimensional, minimizing erasure under a cap on silent corruption, and the joint design a Stackelberg game against a Kullback-Leibler (KL) indistinguishability metric, solved by an alternating convex relaxation and corroborated by an independent reinforcement-learning policy. On a rate-$3/4$ $(16,12)$ parity-check (LDPC) code against an observer who knows her own channel but whose channel the transmitter knows only statistically, shaped litter cuts the observer's Stein detection exponent by $41$ to $48\%$ relative to uniform cover, flat over a $10$~dB span and in the activity rate. Below the point where the legitimate link turns reliable, no cover distribution admits an operating point; at that point shaping costs about one percent of throughput, and nothing measurable $2$~dB above it.

Contextual image misuse pairs an image with a misleading claim. We study image-claim correspondence in fact-checked pairs containing out-of-context reuse, visual manipulation, or both. Existing pair-based detectors often compress the two modalities into a global compatibility score or learn a highly flexible interaction module, which can obscure a decisive local mismatch. We introduce directional multiscale coverage, a compact representation that summarizes local image-claim affinity in both directions and at three spatial scales. At each scale, each direction is summarized by its mean, lower quartile, and two thresholded support ratios; the signed difference between directional means completes a nine-dimensional scale descriptor. Concatenating the three scales yields a compact local representation for a lightweight global-local classifier. Under leakage-aware three-fold, three-seed evaluation on the Snopes subset of the Fauxtography benchmark, UNMATCH achieves 69.82 Macro-F1 and 71.05 balanced accuracy, exceeding the MCOT adaptation by 2.60 and 2.16 points. A matched-reassigned intervention shows that breaking the observed pairing lowers coverage and increases both discrepancy and false-pair probability.

Accurate and rapid prediction of the aerodynamic drag coefficient ($C_D$) is essential for vehicle design, particularly during early-stage design, where many candidate geometries must be evaluated. Although computational fluid dynamics (CFD) provides reliable aerodynamic estimates, its high computational cost limits large-scale design exploration. This paper proposes the hierarchical graph-pooling Transolver (HGPTrans), which combines hierarchical graph pooling with Transolver-based attention to directly predict $C_D$ from vehicle surface meshes. Motivated by the fact that vehicle aerodynamics depends on both local geometric features and long-range interactions among spatially distant surface regions, HGPTrans integrates three complementary components. Graph isomorphism convolutions encode discriminative local geometry, Transolver-style slice attention captures global interactions with linear computational complexity, and information-redundancy-aware hierarchical pooling progressively removes redundant nodes while preserving informative geometric structures. The model is trained and evaluated on the large-scale DrivAerNet and DrivAerNet++ datasets, where it achieves the lowest mean absolute error and mean squared error among the evaluated baselines. Its generalization capability is further assessed through transfer learning on a real-vehicle dataset containing both sedans and sport utility vehicles (SUVs), achieving relative $L_1$ errors of 1.56% (sedans) and 2.12% (SUVs) with an inference time of approximately $0.293$ s per vehicle. This corresponds to an acceleration of several orders of magnitude relative to high-fidelity CFD while keeping the predicted drag coefficients within a few percent of the CFD reference. Ablation studies confirm each component's contribution and reveal the effects of depth and pooling ratio.

Complementarity analyses of multimodal systems commonly report an oracle ceiling (the fraction of examples on which at least one unimodal branch is correct) and interpret the gap between it and realized fusion accuracy as recoverable headroom. We show that this ceiling is not a fusion target. Using frozen self-supervised audio and visual encoders with a trained head, we measure the ceiling and realized gain on CREMA-D under two audio encoders (one whose fine-tuning lineage includes CREMA-D, one clean) and on EAV under the contaminated encoder. Replacing the fine-tuned encoder with a self-supervised one triples the headroom, from +0.055to +0.186. A second clean encoder and a controlled degradation of the contaminated one fall on the same curve, so provenance moves the headroom by changing audio-branch accuracy. Concatenation realizes at most about half of the headroom (converted fraction 0.55), and on EAV the converted fraction is indistinguishable from zero despite larger headroom. A learned router is beaten by plain concatenation everywhere. Oracle headroom measures branch disagreement, not a fusion budget.

Long-context clinical AI systems can miss relevant patient history when prior admissions fall outside the active reasoning context. In ICU monitoring, this can cause early vital-sign drift to appear nonspecific even when it resembles a prior deterioration pattern. SMARtCARE addresses this gap through a four-state clinical decision-support architecture: Stable, Meta-cognitive, Assisted, and Regulated (Revoked). Rather than automatically retrieving prior records, SMARtCARE uses a lossy six-channel fingerprint of the patient's prior trajectory. When current drift matches that fingerprint and the prior record is absent from context, the system raises a Meta-cognitive escalation for clinician review; full retrieval occurs only through clinician action in the Assisted state. A patient-identity guard is designed to enforce correct attribution across data loading, logging, and audit layers. Evaluation combines a synthetic Monte Carlo study that validates the state-transition logic and estimator stability, not clinical performance, with real-data runs on both the MIMIC-III and MIMIC-IV Clinical Database Demos. On MIMIC-III, one prior-pattern recurrence was identified among 14 two-admission patients; on MIMIC-IV, the same pipeline produced no fingerprint matches among 9 two-admission patients, which illustrates a key limitation of a fixed canonical pattern library. Across both runs all logged decisions were fully traceable and correctly attributed. The results support SMARtCARE as a traceable, privacy-aware mechanism for surfacing middle-context risk; they are not a clinical efficacy claim.

Accretion onto magnetically charged black holes in a perfect fluid dark matter (PFDM) background opens a new window for testing strong-field gravity. This paper investigates the radiative properties and optical appearance of a thin accretion disk surrounding a charged-PFDM black hole. We numerically compute the radiative energy flux, temperature distribution, and radiative efficiency, and employ a ray-tracing method to construct the direct images, secondary images, redshift distribution, and observed flux. By comparing with the Schwarzschild and pure PFDM cases, we find that the thin-disk efficiency of the charged-PFDM black hole lies between the two, with the magnetic charge partially counteracting the dark-matter-induced efficiency enhancement; for M87*, the efficiency is estimated to be 7%–8%. The optical appearance is predominantly governed by the PFDM parameter, while the magnetic charge plays only a marginal role. Larger inclination angles give rise to stronger Doppler asymmetry, producing the characteristic “hat-like” shape. These results provide falsifiable predictions for future high-resolution observations, such as those by ngEHT, to distinguish dark-matter environments from magnetic-charge effects.

We prove the Fraser–Sherman-Bennett conjecture on the quasi-equivalence of positroid cluster structures arising from admissibly relabeled plabic graphs. We construct a tilting module from the boundary necklace and obtain a Frobenius categorification of the relabeled cluster structure. By identifying localized characters of rank-one modules with Plücker coordinates, we match the stable images of the mutable face modules with the summands of a reachable cluster-tilting object in the stable category. The induced derived equivalence yields a unimodular transformation of relative index lattices intertwining the full extended exchange matrices. These comparisons show that the identity map on the coordinate ring is a quasi-cluster isomorphism between the two cluster structures.

The para-Bannai–Ito polynomials are orthogonal on a Bannai–Ito bi-lattice that is two-sided — it extends on both sides of a central value $c_0$ — and depend, besides the two grid parameters, on an isospectral deformation parameter. Their Jacobi matrix $J$ is invariant under the reversal of the sites at one value of that parameter only, the persymmetric point, and we determine its continuum limit there. Because the grid is two-sided, the natural object is the shifted operator $Q=J-c_0$, and the limit is taken about the centre of the band: $Q$ contracts to a first–order Dunkl operator, a supercharge, and $H=Q^2$ to the generalized Pöschl–Teller (Scarf I) Hamiltonian, with little $-1$ Jacobi eigenfunctions — supersymmetric quantum mechanics with reflections, the reflection entering the square root of the Hamiltonian. The second grid parameter adds a reflection term to the Hamiltonian itself, and the bi-lattice is, up to a shift, the spectrum of the limiting supercharge. The convergence of the eigenvectors is proved directly on the polynomials. On this grid the persymmetric model transfers states perfectly, under an arithmetic condition on the parameters, but never revives fractionally, revival being the work of the deformation; both properties pass to the limit. These results are for an even number of sites; for an odd number of sites the limit is instead the matrix supersymmetric pair of Pöschl–Teller Hamiltonians, with the reflection in the Hamiltonian as a Dunkl potential. The bispectral pair of the discrete model goes over to the Dunkl realization of the anticommutator algebra of the little $-1$ Jacobi polynomials.

We consider generalised Casimir operators associated with irreducible representations for finite-dimensional, semisimple, complex Lie algebras. Using Klimyk's formula and Okubo's formula, we derive a rational ordinary generating function for the Casimir operator eigenvalues for simple Lie algebras which does not rely on knowledge of the Weyl group. We also directly reproduce generating function expressions for the classical Lie algebras originally due to Perelomov and Popov.

How many assignments can a language model recall before it loses track of which value belongs to which entity? We measure this limit using continuous recall curves for 12 models at or below 3B parameters and a threshold sweep over 30 open models up to 12B. On the continuous curves, the load at which recall falls halfway to chance follows $K_{50}=cN^{\alpha}$, with $\alpha=0.820$ and $R^2=0.73$. The broader sweep shows an eightfold range associated with pretraining recipe, although the continuous curves show no detectable recipe effect after controlling for scale, with few modern models in the fit. We derive why interference can lower measured capacity by reducing single-binding recall even when the load-dependent recall profile is unchanged. Direct task training also exceeds the extrapolated zero-shot law, but different measurement criteria prevent interpreting that comparison as a capacity gain. Its formation times follow a power-law form in two independent codebases, conditional on runs that succeed. Together, these results characterize capacity at the model's query interface. Bounds on joint recall and a decomposition of policy errors connect this measurement to working memory and instruction following, without treating recall as a measure of alignment. The controlled task also provides a baseline for testing whether binding limits constrain world-state tracking; the present experiments do not measure state updates or downstream transfer.

Successful training of deep neural networks is highly dependent on the distribution of the initial weights. If the weights are too large, network training blows up; if they are too small, the model fails to learn features. Stable initialization is the optimal moderation between these two extremes. The conventional theory of random networks uses the Central Limit Theorem to control inter-neuron dependencies, which introduces distributional approximation error and coupling between layers. For networks with sine activations, we derive the uniform-phase initialization, which obviates distributional approximation and fully decouples the layers. Ours is the first work to use the sine function's periodic symmetry. Models trained with the uniform-phase initialization outperform the state of the art in neural representation tasks like image and audio fitting. We find that our untuned models are competitive with the best-tuned baselines from previous work and support $\mu$P width scaling.

We propose a new $q$-derivative operator built directly from Jackson's $q$-number formulation, designed to preserve the structural properties of standard differential calculus while incorporating deformation effects. By analyzing $q$-deformed Heisenberg algebras, we demonstrate that this formulation maintains the consistency of thermodynamic quantities-such as internal energy, particle number, and specific heat-in dilute gas limits without requiring ad-hoc chain-rule modifications. Furthermore, we explore the physical implications of algebraic deformation using the Biedenharn-Macfarlane realization, showing how $q$-deformation induces intrinsic anharmonicity in quantum oscillator spectra and affects multi-level quantum systems ($d \ge 3$). Applying this algebraic scheme to superconducting transmon qubits, we derive analytical pulse-shaping corrections that generalize the Derivative Removal by Adiabatic Gate (DRAG) technique, offering a robust method to suppress computational leakage in ultra-fast quantum logic operations.

Online audio-visual target-speaker extraction aims to remove competing voices while preserving speech quality and bounding lookahead. Existing extractors are built and evaluated for separation on synthetic mixtures, leaving listening quality and meeting behavior largely untested. We introduce Audio-Visual Personalized Voice Quality Enhancement (AV-PVQE), which approaches these requirements from the other direction. We start from a personalized speech enhancement model that reconstructs a requested voice at high quality but confuses the target in 46% of two-speaker mixtures despite clean enrollment. Adding mouth features at its speaker-conditioning input and jointly fine-tuning the visual and reconstruction networks reduces this rate to 1.6%, with no future frames and 20 ms of algorithmic delay. Compared with an online autoregressive audio-visual extractor, AV-PVQE yields separation gains on two synthetic benchmarks and larger gains on recorded meetings, and keeps its advantage on excerpts with more speakers than the fine-tuning mixtures. In personalized P.835 listening tests on two meeting corpora, it improves overall quality over this extractor by 0.57 and 0.63 MOS, with similar mean rating relative to the starting model. Preservation and rejection tests show that it keeps the target intact when no competing voice is present and suppresses competing speech when the target is absent.

We establish sharp temporal lower bounds for the full spatial $L^2$-norm of energy solutions to $\partial_tu-\Delta u=vu$ on $\mathbb{R}^n$, with $n\geq3$. For global solutions, nonvanishing at a single time implies the lower bound $ce^{-Ct}$ when $v$ is small in the scale-invariant space $L^\infty_tL^{n/2}_x$, and $ce^{-Ct^2}$ throughout the subcritical range $v\in L^\infty_tL^p_x$, with $p>n/2$, without smallness. The linear time exponent is optimal already for the free heat equation, while we show that the quadratic exponent is optimal for bounded complex potentials on $\mathbb{R}^3$. No additional regularity of the potential is required. The proof uses a convexified mixed Carleman estimate whose norms are designed together with the final time truncation. Integration over spatial centers removes the heat-kernel weight and yields quantitative comparisons of the full solution norms. To prove quadratic sharpness in the Euclidean energy class, we construct a time-dependent function $g$ that localizes Meshkov's periodic parabolic example while keeping the induced potential bounded across the zeros of the periodic profile.

Mission-critical UAVs increasingly rely on split vision-language model (VLM) perception under tight onboard-resource and wireless-communication constraints. However, corruption of transmitted intermediate features creates a deployment mismatch for clean-trained split interfaces, while stronger channel-aware codecs can impose substantial onboard cost. We present FreshLatent, a lightweight channel-aware latent adapter that trains a power-normalized encoder-decoder through wireless corruption while keeping the surrounding VLM frozen. We formulate deployment around a mission-conditioned perception requirement and embedded interface cost, linking channel quality and communication budget to the operating conditions under which perception remains usable. At 0 dB and the tightest communication budget, FreshLatent improves gIoU and cIoU over clean split compression by 20.79 and 20.87 points, respectively. At the most adverse evaluated SNR (0 dB), across all three communication budgets, FreshLatent recovers 63.5-69.1% of the gIoU improvement achieved by a much heavier, range-trained feature-JSCC codec. On an NVIDIA Jetson AGX Xavier in 10-W mode, FreshLatent uses 37-40x fewer encoder parameters, 7.7-9.9x lower edge-interface latency, and 8.8-10.0x lower edge-interface energy than the heavier codec. Together, these results show that lightweight channel-aware adaptation can recover a substantial fraction of the robustness of a much larger communication interface while broadening quality-valid operation under constrained wireless conditions.

Machine-learning policies have attracted increasing interest for ride-hailing fleet control in recent years. Reinforcement learning, in particular, requires a structured simulation environment that specifies observations, actions, rewards, and decision epochs for training and evaluation. For electric fleets, this environment must also capture the interaction among stochastic demand, vehicle operations, and capacitated charging infrastructure. We present OpenHail, an open-source Gymnasium environment for joint control of electric ride-hailing fleets. Its fixed-size observation–action interface exposes request assignment, repositioning, and charging to a single policy. The event-driven simulator represents requests with pickup deadlines, vehicle job queues, battery dynamics, and finite-capacity charging facilities with first-in–first-out queues. A configurable decision-epoch mechanism separates internal simulator events from policy interactions, supporting event-driven, periodic, hybrid, and policy-requested control within the same operational model. The software provides seeded instances, feasible-action utilities, evaluation tools, operational metrics, and baseline policies. The source code is available at https://github.com/tommaso-schettini/openhail.

Readable annotations are essential for interpreting statistical graphics, yet conventional single-color lines and labels can lose visibility when they cross backgrounds with heterogeneous luminance. We introduce a contrast-aware annotation framework implemented in the ggtwotone R package for ggplot2. The framework combines dual-stroke rendering, adaptive text-color selection, and perceptually guided highlight palettes to improve annotation visibility across light and dark regions. A shared contrast-adjustment engine supports WCAG- and APCA-based contrast criteria, reducing the need for manual color adjustment. The package provides contrast-aware geoms for segments, curves, paths, mathematical functions, regression overlays, and text while remaining compatible with standard ggplot2 workflows. A simulation-based evaluation across heterogeneous background colors demonstrates improved worst-case contrast for dual-stroke annotations and adaptive text selection, while also identifying trade-offs in perceptual separation as the number of requested highlight colors increases. Applications to statistical graphics and scientific images illustrate the framework in practical visualization settings. Together, these tools provide a reproducible approach for incorporating contrast considerations directly into graphical annotation.

We present a frequency-modulated (FM) haptic display based on piezoelectric vibrating actuators. Existing haptic displays commonly encode haptic intensity through the deformation amplitude of individual haptic pixels. Although amplitude-modulated (AM) approaches have enabled compact haptic pixels, independently controlling the deformation amplitude of a large number of pixels can require increasingly complex and bulky driving systems, posing challenges for scaling toward high-density, large-area wearable displays. To address this scaling challenge, we investigate an FM design principle in which haptic intensity is encoded through vibration frequency. We further develop a Shared-Source Frequency Modulation (SSFM) structure in which multiple haptic pixels are powered by a common power amplifier while their vibration spectrum are controlled individually, reducing the need for independent high-power amplification at each pixel. A proof-of-concept piezoelectric haptic display was built and evaluated on rendering spatial and temporal haptic patterns through volunteer tests. The results show that participants reliably distinguished spatial and temporal patterns encoded using FM principles within the investigated operating range. These findings demonstrate the feasibility of FM-based distributed haptic rendering and suggest a potential pathway toward more compact driving architectures for future high-density, large-area wearable haptic displays.

Audio large language models allow users to interact with the model through speech. When an input recording is too degraded, the model may misinterpret the user's query and respond based on an incorrect transcription. In this paper, we study model-conditional transcription reliability: whether an Audio LLM can recognize when its own transcription is unreliable. We first prompt the Audio LLM to assess whether its own transcription would be reliable, and find that the model is a poor judge of its own transcription reliability: in most cases, it predicts that its transcription will be reliable. We find that existing approaches, including speech quality predictors, audio LLM generation uncertainty, and transcript-conditioned WER estimation, provide limited signals for detecting transcription failures. In contrast, we discover that transcription reliability is strongly represented in the model's audio-encoder representations. Based on this observation, we devise a lightweight reliability predictor that operates on representations extracted by the frozen audio encoder and predicts the reliability class before generation. The reliability predictor can trigger a clarification request from the user when their voice query is predicted to be unreliable, while allowing reliable queries to proceed without modifying the underlying Audio LLM. Our predictor achieves 81.10% in-domain and 78.09% cross-domain macro-F1 scores, outperforming the strongest baselines by 10.33 and 11.93 points, respectively. Finally, we show that reliability labels can transfer across Audio LLM families, and that transfer performance is closely related to the alignment of their model-specific reliability boundaries.

The para-Racah polynomials, taken at the persymmetric value of their isospectral deformation parameter, define through their recurrence relation a mirror-symmetric tridiagonal Hamiltonian $J$. Its spectrum is a quadratic bi-lattice, the union of the squares $(s+a)^2$ and $(s+c)^2$ of two interlaced arithmetic progressions of unit step, and it exhibits fractional revival when the grid parameters $a$ and $c$ satisfy a Diophantine condition. As the number of sites grows, $J$ contracts to the trigonometric Pöschl–Teller Hamiltonian on a finite interval. We establish this twice, on the recurrence relation and on the polynomials themselves. Both grid parameters survive in the limit, the difference $c-a$ as a scalar centrifugal term at the interior point fixed by the mirror, and $a+c-1$ as the strength of two confining singular walls at the ends. These walls turn the spectrum into the quadratic pair $(s+a)^2$, $(s+c)^2$, with Jacobi eigenfunctions $P_s^{(a+c-1,\,\mp(c-a))}$, and the two sublattices converge to two families that differ only by the sign of the second Jacobi parameter. The mirror symmetry of $J$ becomes the reflection about the midpoint, which grades the two families and carries the fractional revival to the limit. Algebraically, the bispectral pair of the discrete model, which generates the Racah algebra, becomes a pair of generators of the Jacobi algebra, one rung down the Askey–Wilson hierarchy, which is the dynamical algebra of the Pöschl–Teller Hamiltonian.

Recent generative video editing models enable video content modification (e.g., changing a character) but target short clips. Extending them to full multi-shot videos requires tedious work to locate relevant content across shots, segment it into clips, craft context-aware editing prompts for each clip, and repeatedly articulate complex editing intent. To address this, we explore an interaction paradigm for editing through underlying video structures (e.g., scripts, scenes, characters, shots, and their relationships). We present CraftTrace, an interactive prototype that transforms a video into a malleable, multilevel structure for generative editing. Users work in task-centric workspaces to modify elements or reshape relationships, while an AI agent translates and propagates changes across the video. A user study and expert review show that this structure helps users understand videos, formulate and refine editing intent, and explore alternatives, supporting rapid prototyping during early-stage exploration and full video post-production.

Martingale posteriors and related predictive resampling methods replace the likelihood–prior pair used within Bayesian inference with a predictive model for future observations. These methods are simple to implement and increasingly popular due to their computational efficiency, but little is known about their ability to accurately quantify uncertainty. In this work, we study the concentration and calibration properties of the martingale posterior for general functionals, and show that credible sets can systematically undercover if the predictive algorithms are not carefully tuned. We propose a simple remedy: the bagged martingale posterior. Rather than starting every predictive path from the observed sample, we start the paths from random bootstrap resamples of the data and then amalgamate the resulting draws. Critically, this approach incurs no additional {simulation} cost compared to standard predictive resampling, and delivers conservatively calibrated credible sets. This scalability enables application to a range of challenging examples, including sparse high-dimensional regression and nonparametric conditional quantile models.

Foundation machine learning interatomic potentials (MLIPs) promise near-density functional theory (DFT) accuracy across broad areas of chemistry and materials science. However, their performance in describing systems and processes relevant to heterogeneous catalysis remains underexplored. Here, we introduce HetCat26, a collection of benchmark tests designed to assess pre-trained MLIPs across key aspects of catalytic modeling, including surface energetics, metal-metal oxide interactions, adsorption, and catalytic reaction networks. Evaluating fifteen foundation models, we find that performance on existing general materials benchmarks is only weakly predictive of performance on HetCat26; transferability to heterogeneous catalysis cannot be inferred from these general benchmarks. Across the benchmark tests, current models describe surface energetics and, perhaps surprisingly, reaction barriers well, whereas larger errors are observed for adsorption, and DFT site preferences are often not reproduced. Two models, eSEN-30M-OAM and MACE-MH-1-OMAT, nevertheless achieve high accuracy across the properties evaluated. Beyond model benchmarking, HetCat26 highlights the importance of training data consistency: PBE and PBE+U calculations should not be mixed within a training set. Overall, we identify challenges limiting the transferability of current foundation MLIPs to heterogeneous catalysis and, more broadly, to chemical reactions at interfaces, providing guidance for the development of the next generation of foundation models.

Mobile virtual reality (MVR) provides low-cost access to extended reality (XR), but its limited input restricts use of common locomotion techniques such as head-decoupled and velocity-controlled steering. Using a low-cost controller with an audio-based button, we compared gaze- and controller-directed steering and teleportation in MVR. We included a controller pitch-based speed-control technique enabling continuous head-decoupled steering. Performance and participant feedback indicate that gaze was better suited to teleportation, whereas controller pointing better supports steering in primed search. We compare with similar techniques in standard VR and derive design considerations for low-cost, narrow-FOV hardware, demonstrating the potential to support varied travel techniques and velocity control in low-fidelity XR.