Computation and Language
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Friday 11 September
Clinical decisions are prospective, but clinical language models are often evaluated on retrospective records that reveal the final diagnosis, treatment response, and outcome. Such evaluations may reward the use of future information rather than reasoning under the uncertainty present at the decision point. We introduce a paired benchmark for measuring outcome-conditioned shifts consistent with hindsight bias in clinical temporal reasoning. It contains 171 case reports from the PubMed Central Open Access Subset—40 sepsis and 131 GLP-1/diabetes cases—represented as both textual narratives and human-annotated and LLM-generated textual time series (TTS). For each case, questions are tied to a clinically meaningful cutoff and paired with a prospective reference answer and an outcome-consistent hindsight trap. Models answer each question using either a TTS truncated at the cutoff or the complete timeline; additional conditions vary the narrative source (original or synthetic) and TTS annotation source (human or LLM). We evaluate accuracy (Acc), hindsight trap rate (HTR), answer instability rate (AIR), and hindsight bias rate (HBR), each of which captures different signals of hindsight bias. Across GPT 5.6 Sol, Gemma 4, GLM 5.2, and Opus 5, full timeline exposure produces consistent hindsight-sensitive shifts, while temporal masking reduces bias without lowering accuracy.
Machine Learning
A Machine Learning API for Earth Observation Data Cubes Based on openEO
Earth Observation (EO) data are increasingly organized as spatio-temporal data cubes, while machine learning (ML) methods operate on tabular feature matrices or structured tensor inputs. This mismatch forces platform-specific transformations that are difficult to reproduce or transfer across cloud infrastructures. The openEO specification provides a unified interface for EO data access and processing across heterogeneous backends, but lacks a standardized approach for ML integration. We propose a process-level ML specification for openEO structured into three stages: model initialization, model actions (training, tuning, inference, validation), and model management. It supports classical algorithms such as Random Forest and SVM, as well as deep learning architectures for time-series and spatial patch-based modeling, including TempCNN, Temporal Attention Encoders, and foundation model inference. Three prototype implementations in R and Python demonstrate feasibility across diverse technology stacks. A crop type mapping use case demonstrates cross-backend interoperability by submitting an identical process graph to independent R and Python backends and comparing predictions and evaluation metrics. Two further use cases demonstrate deep learning on time series and foundation model inference, each executed on a dedicated backend. The prototypes reveal, however, that full cross-backend portability requires deeper harmonization of serialization formats and execution semantics than the process level alone can enforce; backend library versions and preprocessing conventions outside the specification's boundary also affect reproducibility. Addressing both through explicit backend conformance profiles represents the most important near-term direction. The specification advances the reproducibility, portability, and accessibility of ML workflows on EO data cubes across cloud platforms.
Read the full abstract
Databases
COMPASS: Steering Distributed Vector Search with Scientific Knowledge Graphs
Vector databases use hashing to partition data across "shards," logical units for distributed execution. This placement, however, destroys semantic locality, forcing each query into scatter-gather limited by the slowest shard. Vector-space clustering can help, but scientific evidence is often connected by factual relations that do not align with embedding distance. We present COMPASS, a framework that uses a knowledge graph (KG) to determine data placement and query-time shard selection. COMPASS detects communities, splits oversized communities, inserts embeddings by subject entity, and routes queries to a small set of shards. Across four biomedical KGs, our method searches only 13-18% of the corpus while preserving broadcast recall and recovering up to 2.6x more multi-hop evidence than an embedding-based baseline. On 15 HPC nodes, COMPASS sustains 7.9x higher throughput with lower tail latency than hash-based broadcast. These results show that KG structure provides a compact complement to embedding geometry for scalable vector search.
Read the full abstract
Plasma Physics
Sensitivities of electron-scale core transport in MAST Upgrade
Integrated modelling of the Mega Ampere Spherical Tokamak Upgrade (MAST-U) indicates that turbulent electron heat transport can dominate over ion transport across a range of operating regimes. Only a limited set of instabilities are able to produce this behaviour, with the primary candidates being microtearing modes (MTMs) and electron temperature gradient (ETG) driven modes. This work investigates electron-scale core transport in two L-mode plasmas and one H-mode plasma on MAST-U using local gyrokinetic analysis. The linear and nonlinear sensitivities of ETG modes are examined, with particular focus on their dependence on electron temperature gradients and ExB shearing rates. In L-mode discharges, ETG modes are found to be linearly unstable over a broad radial region and can drive experimentally relevant levels of electron heat transport, particularly towards the outer core. The transport is strongly sensitive to both the electron temperature gradient and the ExB shear, with nonlinear simulations showing stiff transport and good agreement with experimental estimates within uncertainty. Towards the core, however, ETG-driven transport is reduced and is insufficient to fully explain the observed anomalous heat flux. In contrast, in the H-mode plasma ETG modes are found to be stable or only weakly unstable in the core, resulting in negligible electron-scale transport. This is attributed to higher equilibrium pressure gradients and reduced electron temperature gradients, both of which stabilise ETG turbulence. Following the collapse of the core rotation profile, ETG modes can become unstable towards the outer core, but typically remain insufficient to account for the full level of transport. Overall, ETG turbulence can play a significant role in setting electron heat transport in L-mode plasmas, particularly at larger radii, but is unlikely to dominate in high-performance H-mode conditions.
Read the full abstract
Software Engineering
Usage Control in Public Data Exchange Infrastructures for Future Digital Governance
Public data exchange infrastructures, such as X-Road, are essential to data-intensive digital public services and the EU digital economy. Their next generation requires advanced data usage control (DUC) mechanisms that address emerging requirements, including privacy by design, to mitigate privacy threats inherent in large-scale automated data exchange. Although research within the individual disciplines already addresses many of the emerging requirements, significant practical challenges arise at the intersection of the disciplines. This paper analytically examines the practical challenges and requirements of data usage control in public data exchange infrastructures. We examine the Estonian X-Road system and emerging use cases through an interdisciplinary lens that integrates formal verification and monitoring of usage control, requirements engineering, and digital governance research. We formulate key requirements and propose a conceptual vision for developing usage control in public data-exchange infrastructures to enable future digital-governance systems to comply with privacy requirements by design.
Read the full abstract
Numerical Analysis
MORTIS: Quadrature-compatible cofactor gauges and optimal material margins in finite elasticity
We develop a discrete design theory for cofactor reference forms that preserve the complete finite-element elasticity tangent. A fixed continuous potential generates the coefficient, while the variation space and quadrature determine compatibility. We classify the exact local potential spaces for all-order tensor elements, anisotropic spaces, cubic serendipity and quadratic tetrahedra. At a stress-free Mooney-Rivlin state with a common positive material-coefficient sum, a convex loss expresses the attainable quadrature-point material margin. The optimum is attained; under stated space and sampling assumptions, its ideal value is attained precisely when the physical identity potential is compatible. A conforming fixed-candidate example proves a strict margin loss, linear in curvature amplitude and uniform in cell size, under displacement enrichment. Explicit positive candidates accompany this restriction. A Gram identity gives the sharp constant-coefficient defect at a prescribed margin in the stated norm. Compatible and reference-defect-corrected forms retain the original tangent through a deformation-independent external-boundary Hessian; equal potential traces give equal exact assembled gauges. Quadratic tensor geometry admits a stress-free physical coercivity bound uniform in mesh size and displacement order under explicit material, shape, quadrature and boundary assumptions. Certified curved tetrahedral shear families with zero recovered cell pressure establish finite-deformation regimes, while a loaded pressure-curvature counterexample identifies their limitation. The theory and observations distinguish local exactness, material certificates, complete-reference coercivity and unchanged equations.
Read the full abstract
Systems and Control
From Winding-Fault Geometry to Reliability: Estimation and Prognosis of Stator Inter-Turn Faults
This paper develops an integrated framework for estimation, prognosis, and reliability assessment of stator inter-turn faults in induction motors. The fault is characterized by its severity, defined as the fraction of short-circuited turns, and its spatial orientation. A geometric fault model shows that the resulting output signature is affine in the fault severity and exhibits a fundamental spatial periodicity. Exploiting this structure, an augmented-state particle filter jointly estimates the nonlinear electromechanical state and the unknown fault severity while quantifying posterior uncertainty. The estimated degradation is then propagated using four prognostic models: linear trend, Holt exponential smoothing, Bayesian degradation, and particle-based forecasting. Their predictions are connected to threshold-crossing remaining useful life (RUL), first-passage reliability, degradation-dependent hazard reliability, and a Weibull lifetime benchmark, thereby providing both deterministic and probabilistic health assessments. Numerical results demonstrate accurate online fault estimation and output reconstruction, characterize the effects of degradation pattern and prediction horizon on prognosis, and show consistent reliability and threshold-crossing predictions. Moreover, estimation accuracy remains comparable across the distinct fault orientations induced by the spatial periodicity. The resulting framework provides a unified connection from physics-based inter-turn fault modeling to online diagnosis, degradation prognosis, and reliability assessment.
Read the full abstract
General Relativity and Quantum Cosmology
Effects of the Ekpyrotic Mechanism on Inflationary Phase in mLQC-II
Bouncing cosmological models replace the big bang singularity with a regular bounce, but face a well-known obstacle: the anisotropic shear grows as a^(-6) during contraction, where a is the average expansion factor, threatening to dominate before a homogeneous, isotropic universe can emerge from the bounce. A standard remedy is a scalar field with an ekpyrotic-like potential that turns negative near the bounce, pushing its effective equation of state above unity so it outgrows the shear. Here we examine how this mechanism affects the subsequent inflationary phase within mLQC-II, a modified loop quantum cosmology (LQC) model in which inflation is otherwise generic. Taking a potential combining ekpyrotic and inflationary components, we numerically evolve the dynamical equations across the bounce for various parameter choices. We find that while some choices of the free parameters of the theory allow the ekpyrotic potential to dominate near the bounce and resolve the shear problem, it substantially reshapes the post-bounce evolution. In particular, parameter choices that yield sufficient inflation without the ekpyrotic mechanism now often fail to do so. Sufficiently long inflation remains possible but appears to require fine-tuning. As our results are numerical, a more systematic analysis will be needed to establish their generality.
Read the full abstract
Numerical Analysis
RUPA: Nonlinear volume consistency, constraint geometry and singular penalty limits in finite elements
Volume quadrature can change nonlinear finite-element constraints while preserving their reference-state derivatives. We connect an explicit determinant defect to feasible-set geometry and singular mechanical response. For affine tensor elements of coordinate degree $p\ge3$ with $n\ge p+1$ Gauss points per coordinate, determinant volume is exact precisely when $2n\ge3p$. Below that threshold we construct a boundary-fixed cubic defect at every order. The same directions yield a full-space cube-root residual–distance bound under explicit cell-support, coefficient and physical-norm assumptions, with mesh-uniform upper constants at fixed order. With all cell-pressure equations retained, the volume Jacobian gains rank at nearby feasible states despite agreement through second derivatives at rest; the cube-root exponent is sharp on each fixed mesh. A general localized-minimum theorem shows that the first reduced compatibility term contributes its weighted square to the leading energy in a joint small-load, large-bulk limit. Cubic and quadratic defects therefore produce sextic and quartic terms. The full cubic-element interior space has an exact normal form and sharp local error exponents. Curved quadratic tetrahedra supply the second-order contrast, a sparse rational witness and an exact four-Jacobian volume formula. Finite-strain tensor calculations illustrate normalized response separation, with explicit stationary-point, extreme-bulk and pressure-recovery qualifications. The constructive correction preserves exact cell volumes, so quadrature feasibility remains distinct from physical volume preservation.
Read the full abstract
Computation and Language
Causal Analysis and Mitigation of Spurious Onsets in Full-Duplex Speech LLMs
Speech-to-speech LLMs like Moshi, and its derivative PersonaPlex, can listen and speak concurrently through full-duplex generation. However, they can begin speaking inappropriately during prolonged user silence: under digital-zero input, Moshi and PersonaPlex initiate speech in 30% and 27.5% of five-minute continuations, respectively. What causes this spurious speech? We investigate two hypotheses: either repeated sampling selects speech despite persistently low onset probabilities, or self-conditioning on nonspeech outputs causes an abrupt spike in onset probability. We find that, at every observed onset, speech probability spikes by over nine orders of magnitude in one 80-ms frame, supporting the latter hypothesis. Then, to suppress these onsets without blocking genuine responses, we ask a causal counterfactual question: is the model responding to user speech, or would its next-token distribution remain similar if the preceding user input were muted? Accordingly, we suppress onsets whose distributions change little under this intervention. Under realistic microphone noise, our method suppresses spurious onsets, while preserving genuine responses: one-sided 95% lower confidence bounds are 98.68% and 98.82% for Moshi, and 96.90% and 99.25% for PersonaPlex. Our inference-time method runs in real-time without retraining, with 95th-percentile decision time below 61 ms, within the 80-ms frame budget. Our code is available at https://github.com/KentoNishi/icassp27-spurious-onsets.
Read the full abstract
Astrophysics of Galaxies
Host Galaxy UV Emission and Dust Reddening in Little Red Dots
Little Red Dots (LRDs) have been the subject of extensive analysis to place them within the context of galaxy and black hole evolution. Despite several proposed scenarios for the nature of the rest optical emission in LRDs, our understanding of the UV emission remains a matter of debate. Here, we aim to study the origin of the UV emission in LRDs. We conduct emission-line and continuum fitting analysis of thirteen LRDs using a combination of NIRSpec PRISM and medium-resolution G235M and G395M spectroscopy. We isolate and analyze the narrow-line emission arising from larger physical scales in LRDs. We observe that the narrow-line Balmer decrements are typically greater than those predicted for case B recombination and broadly correlated with both the UV slope and the break strength, providing clear evidence that the UV emission originates from galaxy scales and its slope is significantly affected by dust reddening. SED fitting analysis further shows that the UV continuum of LRDs is consistent with a dust attenuated stellar template. Finally, we show through a comparison of the UV slope and the Balmer break strength, that LRDs can be further divided based on break strength, such that the canonical LRD spectral shape (strong break strength ($>$ 3) and flat rest UV continuum) represents the most extreme sub-population. Our results imply that for most LRDs, the UV emission originates primarily from the young host galaxy, and that the characteristic optical emission may be due to a combination of dust reddening and intrinsically red accretion processes.
Read the full abstract
Machine Learning
Learning to Solve Hard Problems in RL for LLMs by Never Giving Up
We demonstrate that training LLMs with RL does not improve performance equally across a dataset. RL shows large improvements on easy problems that an LLM is already good at solving, but small improvements on hard problems. We call this the Matthew Effect in RL for LLMs, after the phenomenon of cumulative advantage from economics and network science summarized as "the rich get richer". The naive explanation is that hard problems require more compute to find a solution. We argue that modern RL methods are exacerbating the issue by wasting too much compute on easy problems and instead should dynamically reallocate how they use compute. We introduce Never Give Up (NGU), a simple adaptive sampling method that keeps generating samples for a problem until one is correct. By leveraging asynchronous RL, this naturally uses fewer samples to filter out easy problems and allocates more compute to solving harder problems. We investigate the design choices that affect NGU, such as off-policy robustness, and develop a set of best practices. On the math benchmark Deepscaler, NGU improves performance per compute, especially on harder problems. On a recent coding task, Manufactoria, standard GRPO with a per-test reward fails to fully solve problems that have a range of easy and difficult tests. NGU iteratively improves, solving harder and harder tests, until it learns to fully solve coding problems.
Read the full abstract
Fluid Dynamics
Large-scale dynamics of equatorial thermal spots
For a model of a thin spherical layer of an incompressible rotating fluid subjected to the action of Coriolis and buoyancy forces, both analytical solutions in the form of stationary thermal spots moving along the equator and self-similar solutions are found. The influence of the parameters determining the shape of the spots on their propagation speed and direction is studied. It is found that the speed and direction of motion of the spots depend not only on the sign of their thermal contrast with the background flow, but also on the ratio of the semi-axes determining their shape. In particular, if the ratio of the equatorial semi-axis to the meridional one $b/a$ is equal to $\pi/2$, the spots are at rest. If the spots are elongated along the equatorial axis to such an extent that $b/a>\pi/2$, then “cold” spots move westward and “hot” spots move eastward. Under the condition $0.747\leq b/a<\pi/2$, the situation is reversed: “hot” spots move westward and “cold” spots move eastward. Self-similar solutions in the model are realized strictly in the form of circular thermal spots, and their radius varies according to a power law with a scaling exponent $k$, which is determined by the behavior of the cross-frontal buoyancy gradient on the spot contour. We note that the regime $k=1/4$ is realized under the assumption of a constant buoyancy gradient, while the regime $k=1/6$ corresponds to the conservation of total buoyancy.
Read the full abstract
Statistical Methodology
Temporal Fourier Likelihoods with Spatial Hilbert-Space Gaussian Process Approximations
Reconstructing stationary space-time Gaussian processes at unobserved locations is costly when many sites share regular temporal records. We develop a spectral likelihood combining a temporal discrete Fourier transform (DFT) with a Hilbert-space Gaussian process (HSGP) representation of frequency-specific spatial covariance. We derive the exact covariance of the finite-record DFT coefficients and use a Whittle likelihood that approximates distinct frequencies as independent spatial problems. At each temporal frequency, HSGP approximates spatial covariance by evaluating the sampled spectral multiplier at retained Laplacian eigenfrequencies. For models specified by a joint spectral density whose half-spectrum lacks a convenient closed form, this construction avoids repeated Fourier inversion. The fixed spatial basis also permits cached feature projections to be reused in likelihood fitting and held-site reconstruction, with approximation accuracy depending on domain extension and basis size. In a simulation study of such a model, HSGP achieved reconstruction accuracy comparable to a high-accuracy quadrature reference while reducing mean fitting time by 69%. Additional applications to wind-field reconstruction tasks examine the effects of basis rank, temporal record length, and separability, and demonstrate accurate inference on held-out test sites when sufficiently rich bases are used. Taken together, the results indicate that computational savings are attainable when the spectral multiplier can be evaluated directly, numerical spatial inversion is costly, and an adequate basis has rank lower than the number of fitting sites.
Read the full abstract
Machine Learning
Efficient Online Inverse Optimization with $O(d)$ Regret
We give a deterministic algorithm for online inverse linear optimization with regret $O(d)$, uniform in the horizon and $O(d^{2})$ time per round. A bound of this order was obtained recently by Dewasurendra, settling a question of Gollapudi et al.\ and of Oki and Sakaue, but by an improper rule that enumerates covers at every scale and costs $T^{\Theta(d)}$ a round; ours is the first efficient such bound and the first proper one. We build on the variable-metric framework of Sakaue et al., adding a self-normalized rank-one update, and we replace the $\log\det$ potential by the trace power $\tr(H^{-1/2})$, which is bounded outright and removes the $\ln T$. The bound also holds against an expert that does not optimize, and we give corruption-robust and rank-adaptive variants, and an application to convex minimization.
Read the full abstract
Machine Learning
Certifiably Interpretable Training of ReLU-MLPs for Boolean Tasks with Guaranteed Truth-Table Generalization
As compute scales, models evolve, and training algorithms advance, our ability to explain the increasingly powerful AI systems they enable is eroding. To help safeguard interpretability, we introduce a specialized training algorithm (MACCHIATO) that jointly constructs (i) an explicitly structured $\operatorname{ReLU}$-MLP from partial truth-table observations and (ii) an explicit Boolean circuit over signed literals with $\{\operatorname{AND},\operatorname{OR},\operatorname{XOR}\}$ gates certifying what its subnetworks compute and how they compose. Intuitively, we iteratively project the residuals of a Boolean function onto low-dimensional $\{\operatorname{AND},\operatorname{OR},\operatorname{XOR}\}$-circuit classes and exactly compile the resulting circuit into a $\operatorname{ReLU}$-MLP; we combine $\operatorname{ReLU}$-MLP circuit compilation, ESPRESSO logic minimization, and influence-based variable selection. Roughly speaking, our interpretability certificate is complemented by a statistical guarantee: under the theorem's influence-recovery conditions, if each of the $m$ stage-wise residuals depends on at most $\log_2(B)$ bits, a sample-splitting variant of our algorithm trained on $T$ observations returns a six-layer $\operatorname{ReLU}$-MLP (counting the input layer) of width $\mathcal{O}(mB)$ with truth-table error $\mathcal{O}\bigl(\sqrt{m(B+\log(m/\delta))/T}\bigr)$. On synthetic random-junta tasks, our networks outperform depth- and hidden-width-matched Adam-trained MLPs in several data-sparse or projection-aligned regimes, while the trained ReLU-MLPs are stronger in others. Moreover, in our explicit PyEDA truth-table implementation, the iterative procedure completes in regimes where flat ambient-dimensional ESPRESSO exceeds the three-hour computational budget.
Read the full abstract
High Energy Physics (Phenomenology)
Semi-visible Jets from Sneaky Dark Matter
In this paper, we compare two different ways of generating semi-visible jet signature in non-resonant production mode, and flag the potential advantages of the new generation approach we propose. In addition, we suggest a more efficient matrix element to parton shower matching configuration for this hadronic final state signature. Finally, we perform a reinterpretation study of the ATLAS search for t-channel semi-visible jet production and obtain collider limits on the sneaky DM scenario.
Read the full abstract
Artificial Intelligence
LabAgent: Customize Any Research Hubs for Scientific Discoveries Using AI Agents
Scientific research is a continuous process that emphasizes inheritance. Methods developed by predecessors are often expanded upon by new researchers to explore more novel and in-depth scientific questions. However, the change of lab staff, such as student graduation, leads to a lack of personnel capable of replicating methods. Methods that have been developed with significant effort and resources cannot be continued. To address these limitations, we propose LabAgent, a reproduce and discovery harness tailored for a lab's continuous work. LabAgent employs two mechanisms to guarantee that all skills can be executed and verified and to record the corrective methods and experiences, allowing for direct correction or avoidance of similar errors. We applied LabAgent to drug property prediction, biomedical problem analysis, protein variant effect prediction, and statistical genetics in life science domains. LabAgent ranks first over commercial generalist agents in every domain, and demonstrates accurate reproduction of a published figure. Overall, these results demonstrate that LabAgent can effectively integrate and reasonably expand laboratory knowledge.
Read the full abstract
Artificial Intelligence
Toward Self-Adaptive Physical AI: Can LLM Agents Manage Long-Horizon Physical Tasks?
Large Language Model (LLM) agents offer a promising path toward autonomously managing long-term physical tasks without human intervention. However, physical tasks require agents to continuously observe the environment, make consequential actions, and remain effective as the environment changes. Existing approaches either require substantial data and retraining, or primarily focus on agents operating in the virtual world. In this work, we explore the feasibility of building a self-adaptive physical AI agent that manages long-term physical tasks in a zero-shot manner and adapts to environmental changes without human intervention. We design a multi-agent framework that integrates planning, tool calling, observation, and verification, and evaluate it on agricultural tasks against reinforcement learning (RL) agents under different weather patterns. Our results show that zero-shot LLM agents can achieve comparable management outcomes to RL agents under the same weather pattern and adapt more effectively than RL when evaluated under a shifted environment, highlighting a promising path toward self-adaptive physical AI agents.
Read the full abstract
Number Theory
Arithmetic Siegel-Weil for the spherical Hecke algebra
We formulate a conjectural generalization of the arithmetic Siegel-Weil formula, relating intersection multiplicities of cycles on unitary Rapoport-Zink spaces to central derivatives of local Whittaker functions. The new aspect of the conjecture is that it incorporates the action of certain Hecke correspondences on the Rapoport-Zink space. We verify the conjecture in a low-dimensional case. A significant portion of the paper is devoted to generalizing the classical theory of representation densities for local Hermitian spaces, as this is needed for explicit computation of the relevant Whittaker functions. We expect this generalization to be of independent interest.
Read the full abstract
Mathematical Physics
Microscopic Parametric Correlations and Spectral Rigidity in the Non-Hermitian Threefold Way
We study microscopic spectral correlations between two nearby parameter values in the three non-Hermitian Gaussian bulk classes: $A$ (unconstrained), $AI^\dagger$ (complex symmetric), and $AII^\dagger$ (complex self-dual, with each degenerate doublet counted once). Parameter dependence is modelled by stationary matrix Ornstein–Uhlenbeck evolution, whose microscopic time scale is $N^{-1}$. For every positive integer replica number $n$, we derive exact finite-$N$ auxiliary-field integral representations for joint characteristic-polynomial moments and obtain their bulk asymptotics as $N\to\infty$ with $n$ fixed. With the spectrum normalized to the unit disk, correlations near a bulk point $z_0$, between spectral positions $z_1,z_2$ at times $t_1,t_2$, depend only on $s=N|z_1-z_2|^2+N(1-|z_0|^2)|t_1-t_2|$. Adopting the static Hermitian/non-Hermitian replica continuation yields explicit two-time kernels; for $AI^\dagger$ and $AII^\dagger$ these are replica conjectures and, to our knowledge, the first microscopic two-time kernels proposed for these classes. They determine density correlations, cross-time eigenvalue-count covariances, and equal-time number variances. For a disk of mean eigenvalue count $y$, the latter obey $\mbox{Var}\mathcal N_X(D_y)=\kappa_X\sqrt y+ \beta_X/\sqrt y+o(y^{-1/2})$, with explicitly given constants for each class. Exact perturbation theory further identifies eigenvector nonorthogonality as the mechanism of short-time spectral diffusion and yields a basis-independent condition number for a two-dimensional Kramers eigenspace in class $AII^\dagger$. Direct simulations agree quantitatively with its conjectured inverse-gamma law.
Read the full abstract
Solar and Stellar Astrophysics
Sinking Silicates I: Characterizing the benchmark system containing the T0 brown dwarf CWISE J210640.16+250729.0 using JWST
In this study, we present the full (97.3% complete) 0.8–12.5 $\mu$m spectral energy distribution (SED) of an L/T transition object, CWISE J210640.16+250729.0 (CW2106), using the James Webb Space Telescope (JWST). We provide a full characterization of the host star's elemental abundances and age. We empirically derive the bolometric luminosity ($L_{\rm bol}\approx-4.825$ $\textup{L}_\odot$) of CW2106, and obtain estimates of its mass (M$\approx50-62$ M$_{\rm Jup}$), radius (R$\approx0.83-0.87$ R$_{\rm Jup}$), effective temperature ($T_{\rm eff}$$\approx1213$ K), and surface gravity ($\log~g$$\approx5.28$ dex). We find the near-infrared (near-IR) spectrum ($0.8-2.5 ~\mu$m) is best reproduced with cloudy atmospheric models while the mid-infrared (mid-IR) spectrum ($5-12.5 ~\mu$m) is best reproduced with cloudless models. This suggests a cloud layer restricted to only the deepest observable parts of the atmosphere and is qualified by the lack of a 9 $\mu$m silicate feature. Making use of the Mg/Si ratio of the primary, alongside thermochemical models, we predict the clouds in CW2106 to be composed primarily of enstatite (MgSiO$_3$), removing $\sim23\%$ of the bulk oxygen out of the atmosphere. Future retrieval studies will be able to help investigate the existence and full impact of these cloud species.
Read the full abstract
Disordered Systems and Neural Networks
Fuzzy Erdos-Renyi random graphs: Spectral statistics
Within a random-matrix-theory (RMT) approach, we perform a detailed numerical study of the spectral statistics of fuzzy Erdos-Renyi (ER) random graphs. Specifically, we consider the membership values of the vertices and edges of a fuzzy ER graph model as random variables. In this way, the corresponding adjacency matrices become sparse matrices with random entries. Then, we apply standard RMT measures to characterize the eigenvalues and eigenvectors of ensembles of Hermitian and non-Hermitian adjacency matrices of both undirected and directed fuzzy graphs. We demonstrate that the average degree of the ER model serves as the scaling parameter of the spectral properties of the corresponding fuzzy setups.
Read the full abstract
Number Theory
Dynamical uniform boundedness for unicritical polynomials
We prove the dynamical uniform boundedness conjecture for unicritical polynomials $z^d + c$ over every number field $K$ when $d \geq 4$, and over every number field $K$ not containing $\mathbb{Q}(\sqrt{-3})$ when $d = 3$. Furthermore, our unconditional bounds on perodic points are effectively computable. In the remaining cases, we show that the dynamical uniform boundedness conjecture is implied by the boundedness of Mordell-Weil ranks over $K$ for a specific family of elliptic curves when $d = 3$ and for Jacobians of dynatomic curves when $d = 2$.
Read the full abstract
Graphics
A Sharp Conservative Offset for Marching Cubes
Given a grid-sampled 1-Lipschitz function $f:\mathbb{R}^3 \rightarrow \mathbb{R}$ (such as a signed distance function) sampled on a regular grid, we determine the sharp infimum $\sigma_\star$ such that for every $\sigma > \sigma_\star$ the true zero-level set $f^{-1}(0)$ is contained in the marching cubes mesh for $f = \sigma$. For grids with spacing $h$, the sharp infimum is $\sigma_\star = \frac{\sqrt{3}}{2} h$. This result is proven despite the planar marching cube faces potentially lying on either side of the level set of the trilinearly interpolated approximation of $f$ (which shares the same infimum). The proof instead constructs a convex lift of the corner samples and shows that the high-side portion of each marching-cubes cell lies in the projection of its $\sigma$-superlevel portion, which cannot contain a point of $f^{-1}(0)$.
Read the full abstract
Number Theory
Theta functions after Weil
This is a mostly expository paper on the construction of theta functions on the Siegel and Hermitian half-spaces using the machinery of the Weil representation, especially theta functions weighted by polynomials. When the polynomial in question satisfies suitable homogeneity and harmonicity conditions, the resulting theta function is holomorphic. Non-holomorphic Siegel and Hermitian theta functions can be constructed in a similar way, using a more general class of polynomials.
Read the full abstract
Robotics
Chance-Constrained Belief-Space Maneuver Planning for Autonomous Collision Avoidance Under Uncertainty
Increasing conjunction frequency in low Earth orbit places growing pressure on spacecraft operators to determine not only whether an encounter requires mitigation, but whether sufficient information is available to commit to a maneuver. This work formulates this information-action tradeoff as a belief-space planning problem for conjunctions between a maneuverable spacecraft and an unmaneuverable secondary object. The planner represents the uncertain orbital states as Gaussian beliefs and uses a chance-constrained belief-space Monte Carlo tree search framework to reason over possible future tracking updates before time of closest approach (TCA). A terminal chance constraint limits the probability of reaching TCA above a prescribed collision-risk threshold, allowing the planner to wait for informative tracking while intervening when deferral becomes too risky. We evaluate the approach on eight historical conjunctions from NASA's Conjunction Assessment Risk Analysis dataset. By varying the secondary-object measurement quality and tracking cadence, we generate a total of 96 distinct evaluation scenarios. Across the evaluated conditions, the planner reaches TCA without maneuvering in approximately 40% of episodes while maintaining no terminal collision-risk violations. In contrast, fixed-time rule-based maneuver policies resolve more encounters without maneuvering when intervention is deferred closer to TCA, but at the expense of increasing terminal risk violations. The fraction of episodes reaching TCA without maneuvering depends strongly on tracking quality and measurement cadence, ranging from 76% under accurate, frequent measurements to approximately 18%-20% under the poorest tracking conditions. These results show that tracking quality and frequency are not only inputs to collision-risk estimation: they can determine when intervention becomes necessary.
Read the full abstract
Materials Science
Precessional modes of phonon angular momentum
Magnons are collective precessional excitations of electron spins and are ubiquitous in emerging information technologies. Here, we show that the crystal lattice can support an analogous precessional mode of phonon angular momentum. Superposition of a circularly polarized phonon and a linearly polarized phonon normal to the circular-motion plane generates a transverse angular-momentum component that rotates at their difference frequency and decays on a timescale set by the participating phonon linewidths, defining a mode with its own resonance and lifetime. Coupling between spin and phonon angular momentum hybridizes this precessional mode with a magnon near resonance. Our results establish phonon angular momentum precessional modes as lattice analogues of magnons and identify their spectroscopic and time-domain signatures.
Read the full abstract
Superconductivity
Spatial instability of a homogeneous superconductor after an interaction quench
We investigate the quantum dynamics of a conventional superconductor, described by the attrac- tive Hubbard model, driven far from equilibrium by quenching the Hubbard attraction. We evolve the full density matrix of the system in time using the Heisenberg equation of motion, tracking the spatiotemporal dynamics of the pairing and density fields. We use very weakly inhomogeneous initial conditions to probe the stability of the homogeneous dynamics to spatial perturbations. At initial times after the quench the pairing and density fields remain homogeneous in space while fluctuating in time. However, depending on the quench strength the spatially uniform dynamics can become unstable beyond a threshold time, with emergence of spatial textures. The onset time for the instability depends on the strength of the quench and the initial inhomogeneity. At long times the system evolves towards a state with significantly reduced spatial heterogeneity. The time dependence of the global superconducting order parameter suggests two broad categories of re- sponse to a quench. One of these involves strongly inhomogeneous pairing and bond currents in the nonequilibrium superconductor.
Read the full abstract
Machine Learning
ReCAST: Reward Credit Assignment across Timesteps for Online Diffusion Reinforcement
Training diffusion models with multiple rewards requires distinguishing user preference from reward informativeness. User preference determines how much each reward should contribute to the overall objective; reward informativeness determines when its feedback is useful during denoising. Some rewards can meaningfully evaluate a sample as soon as global structure emerges, but others become informative only when the sample is nearly clean. To address both questions jointly, we propose ReCAST (Reward Credit ASsignment across T}imesteps), the first method, to our knowledge, for per-reward, timestep-dependent credit assignment in diffusion reward fine-tuning. ReCAST separates user preferences from temporal allocation through a reward-by-timestep weight matrix $W$, whose row sums match the user-specified reward budgets $\lambda$, while its column sums are equal, assigning the same total weight to each denoising step. Under these marginal constraints, ReCAST allocates weight according to each reward's informativeness, quantified by its Rényi discriminability gain at each step. These gains telescope to the total discriminability between the reward-induced positive policy and the current policy, providing a basis for temporal credit assignment. We evaluate ReCAST by training SD3.5-Medium under two distinct four-reward settings, each across five reward budgets $\lambda$. ReCAST improves the training rewards in one setting and matches them in the other, improves every held-out judge in both, and is preferred by an independent LLM-as-a-Judge. Together, these results show that ReCAST yields improvements that generalize beyond the training rewards and support its core principle: assigning each reward greater weight at the denoising timesteps where its feedback is most informative.