Information Theory
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Tuesday 1 September
AI-native communication increasingly aims to support prediction rather than reproduce every detail of the source. This shift raises a basic question left implicit by conventional source coding: what should be preserved when the terminal goal is prediction? We take the source-induced predictive state as the fidelity object. It is the distribution of the specified future conditioned on the source observation and shared context. We show that this state is sufficient and minimal for exact predictive preservation. The terminal prediction loss then defines communication distortion as lost predictive performance rather than source reconstruction error. Under logarithmic loss, this distortion equals the conditional mutual information lost through communication. Using the receiver-side predictive state as a Bayes reference, we separate AI-receiver error into predictive value lost in communication, receiver-available value unusable by the model family, and family capability not realized by the deployed model. The same predictive state also suffices for matched compression. For finite-alphabet memoryless sources, access to the raw source gives no rate-distortion advantage over coding the state directly. Applying the same target-conditioned construction to sequential prediction reveals a dynamic boundary. A state induced by a fixed horizon is minimal for that horizon but may not support recursive updating as the target window shifts. Taking the entire future as the target yields a minimal full-future state that updates recursively and admits a Markov representation. Together, these results shift AI communication from source reconstruction to predictive-state preservation.
Optics
Design and Physical Constraints of Synthetic-Frequency Photonic Switching Fabrics
Electro-optic frequency conversion and synthetic-frequency coupling are established functions in integrated photonic devices. Their role within a multiport switching fabric, however, depends on how simultaneous optical connections share spatial paths, frequency channels, and device controls. Here, we investigate how coherent coupling among frequency modes can be incorporated into photonic switching fabrics and identify the corresponding architectural and physical constraints. We show that synthetic-frequency coupling does not increase the number of simultaneous orthogonal frequency channels when all channels are freely accessible, but can establish connections that are otherwise blocked by fixed input frequencies, channel-continuity requirements, or unavailable output channels. Under the tested conditions, coupling over the first three frequency spacings in an $8\times8$ fabric with eight frequency channels per port achieves 96.1% of the blocking reduction obtained with unrestricted inter-mode coupling. We further show that a separate frequency-only conversion stage cannot replace missing spatial connectivity. A nominal reduction in spatial switching elements instead requires a joint element whose spatial state can be programmed independently for each frequency channel. Finally, we evaluate a thin-film lithium niobate resonator model using reported electro-optic coupling and photon-decay scales within a multistage Mach-Zehnder interferometer switching fabric. These results clarify the architectural role of synthetic-frequency coupling and the device-level requirements for incorporating it into integrated photonic switching fabrics.
Read the full abstract
Machine Learning
Scaled Idempotence in Transformer Attention: Paired OV Geometry and Shared-Value Algebras
We identify a recurrent algebraic regularity in Transformer attention: a sparse subset of effective OV operators $T=OV^\top$ nearly closes under composition, $T^2\approx\alpha T$. Across six pretrained endpoints spanning 2.8B–235B parameters, 3.98–8.00% of heads reach squared closure alignment $\mathcal{P}\geq0.9$, while no matched within-layer O/V mismatch does. An exact principal-coordinate factorization, $T=Q_OKQ_V^\top$ and $T^2=Q_O(KDK)Q_V^\top$, separates within-support transport from read–write return geometry. Across all 7,304 heads in nine MHA/GQA models, scrambling only the orientation of $K$ while preserving singular values, norms, factor spans, and principal angles reduces median closure from 0.336 to $1.04\times10^{-4}$; trained orientation wins for 98.64% of heads and in every layer. Constructive searches show that high closure is feasible in every surveyed layer, but usually not attained. Retrospective trajectories in three independently trained lineages further separate broadly available capacity from the orientations attained by final strong heads. Under exact value sharing, headwise closure extends to a right-action algebra, $T_iT_j=\alpha_jT_i$. Seven-model experiments verify the approximate law and reveal distinct oblique projections with a shared value-defined kernel. These results characterize scaled idempotence as a sparse trained orientation within broadly available geometric capacity and show how value sharing extends a headwise relation into a local operator algebra.
Read the full abstract
Number Theory
Nonabelian Chabauty for the Thrice-punctured Line over Cyclotomic Fields
In this paper we study the motivic Chabauty–Kim method, which aims to determine the set of $S$-integral points of $\mathbb{P}^1\smallsetminus \{0,1,\infty\}$, over cyclotomic fields. We focus on the case $K=\mathbb{Q}(\zeta_8)$ and $S=\left\{(1-\zeta_8)\right\}$, where we obtain explicit polylogarithmic Kim functions up to depth $4$ and verify Kim's Conjecture for several primes. We also observe and explain that the Chabauty–Kim locus for the polylogarithmic quotient contains, in addition to the $S$-integral points, certain exceptional points arising from roots of unity in $\mathbb{Q}_p$.
Read the full abstract
Statistical Mechanics
Re-entrance or monotonic: How chemotaxis controls pattern formation
We investigate pattern formation in generic two-species reaction-diffusion systems with mutual chemotaxis. We show that tuning the chemotaxis strength can give rise to striking nonmonotonic variation of patterns, opening the possibility of re-entrant transitions between patterned and spatially uniform steady states. In certain regimes, chemotaxis also induces additional heretofore unexplored instabilities. Furthermore, varying the strength of chemotaxis can drive morphological transitions between spot and stripe patterns.
Read the full abstract
Machine Learning
How Evaluation Choices Shape the Measured Benefit of Online Adaptation in Time-Series Forecasting: Warmup, Learning-Rate Selection, and Adaptation-State Memory
Online adaptation can help edge time-series forecasting under distribution drift, but its measured benefit is sensitive to evaluation choices. We study six public multivariate streams, including building-sensor and smart-meter data, under a leakage-free streaming protocol. We identify two sources of comparison bias. First, the warmup budget of the static baseline has a two-sided effect on seed-mean curves: insufficient warmup undertrains the baseline, whereas excessive warmup can degrade its pre-drift generalization. Across six dataset-backbone settings, the estimated adaptation benefit changes by 3.0 to 18.8 percentage points (pp) over the 1,000-20,000-step warmup range. Second, comparing SGD with momentum (SGD+m) and Adam at a shared default learning rate conflates optimizer quality with rate sensitivity. We select both the warmup budget and each optimizer's online rate using a held-out pre-drift validation slice without accessing test data. Under this procedure, Adam outperforms SGD+m in 310 of 360 cells, and only 4 of its 360 cells fall below the static baseline. Third, we characterize accuracy against adaptation-state memory and A100-measured per-update latency for full, head-only, and calibration-based adaptation. In the evaluated PatchTST frontier settings, several parameter-efficient variants are nondominated on the adaptation-state-memory axis. These findings support a validation-only commissioning procedure. Code, data, and all reported numbers: https://github.com/keiotakmin/tsf-edge-adaptation
Read the full abstract
Cosmology and Nongalactic Astrophysics
Quantifying Sky Map Resolution Requirements for Beam Chromaticity Correction in Sky-Averaged 21 cm Experiments
The 21 cm hyperfine transition of neutral hydrogen provides one of the few direct probes of the early cosmic history. High-redshift detections of this signal could shine light on the poorly understood epochs of the Dark Ages and Cosmic Dawn, periods that remain among the least understood in the Universe's history. However, this signal is masked by bright foregrounds and is distorted by the chromaticity of the antenna used to detect it. Correcting for the chromaticity requires an accurate representation of the radio sky across relevant frequencies. However, base sky maps used for this correction, such as instances of the Global Sky Model (GSM), are limited by resolution, calibration and extrapolation uncertainties, especially when scaled to lower frequencies relevant for Cosmic Dawn studies. This work quantifies how accurately the base map must represent the true sky to produce a reliable beam correction. Using simulated sky data, we generate beam chromaticity corrections generated from progressively degraded versions of the base map, and compare how well they do in comparison to a correction based on the full-resolution map. We find that degradation in map resolution of $\sim 1^\circ$ does not introduce measurable increases in the amplitude of the residuals when correcting for the chromaticity and subtracting three different foreground models (two polynomials and a power law expansion). These results provide practical constraints on the required resolution and accuracy of sky models used in beam correction pipelines, informing future design and calibration strategies for global 21 cm experiments, and future efforts to map the low frequency sky.
Read the full abstract
Instrumentation and Detectors
Helium-Gas-Cooled Cryogenic Current Comparator Integrated with a Quantum Resistance Standard
We report a cryogen-free cryogenic current comparator (CCC) system for precision resistance measurements. A helium-gas chamber was developed and installed on the 4 K stage of a cryogen-free dilution refrigerator equipped with a pulse-tube cryocooler. The CCC probe was housed in this chamber and was cooled through helium gas serving as a heat exchange medium. The metrological performance of this helium-gas-cooled CCC was evaluated through precision resistance-ratio measurements and found to be comparable to that obtained under liquid-helium cooling. A quantum Hall resistance (QHR) device was also integrated into the same refrigerator, enabling QHR/100~$\Omega$ resistance-ratio measurements. The type-A uncertainty reached the 1~n$\Omega$/$\Omega$ level within an averaging time of 100~s, and the resulting resistance-ratio measurements agreed well with conventional liquid-helium-based measurements at the level of a few n$\Omega$/$\Omega$. The system provides this level of measurement performance while consuming less than 1~L of helium gas per thermal cycle.
Read the full abstract
Computer Vision
P-PatchDiff: Progressive Patch Diffusion Models for Low-light Image Enhancement
Recent advancements in low-light image enhancement have leveraged diffusion models for their strong ability to generate perceptually realistic, detailed images. Patch diffusion models further offer a promising solution to size-agnostic image restoration while improving efficiency. However, existing methods typically rely on small, fixed patches (e.g., 64$\times$64) that cannot capture image-level brightness context, whereas enlarging the receptive field improves brightness and colour estimation but substantially increases computational cost. Moreover, low-light images often exhibit uneven brightness across regions, making it necessary to ensure that locally enhanced patches remain visually coherent when combined into the full image. To address these limitations, we propose P-PatchDiff, a scalable progressive patch diffusion framework for low-light image enhancement that dynamically adjusts patch size throughout the denoising process, enabling a gradual shift from local to global views. A Multi-Patch Alignment strategy is also introduced to normalise features across varying patch scales using an estimated global brightness proxy. Rather than pursuing pixel-level reconstruction accuracy, P-PatchDiff focuses on scalability and coherent brightness across the whole image, allowing the model to perceive multi-scale information and better enhance regions with varying brightness. We empirically demonstrate that P-PatchDiff effectively enhances images ranging from 400 $\times$ 600 to 4K and is 80$\times$ faster than existing patch diffusion models while using less than 9GB of memory. The code is available at https://github.com/RuoyuGuo/P-PatchDiff.
Read the full abstract
Optimization and Control
Optimal Gradient-Norm Minimization in Non-Euclidean Hölder-Smooth Convex Optimization
Minimizing gradients of a convex function is an important problem across optimization and learning tasks. The gradient provides a directly computable certificate of approximate stationarity, and its minimization usually implies stronger results than those for minimization of function values. In this work, we study gradient-norm minimization for convex functions that are $(L,\kappa)$-Hölder smooth with respect to the $\ell_p$-norms, $p \geq 1$. We develop algorithms that achieve near-optimal gradient-oracle complexity for this problem. In the smooth case, our results resolve the previously open setting $p>2$. For Hölder-smooth objectives, we close the complexity gap throughout the full $p$-range, including to the best of our knowledge, a gap in the Euclidean case. We provide two families of algorithms: the first one comes with a simple iteration and generalizes a phenomenon known as mirror duality, exploiting dual behaviours of algorithms with errors and inexact computations. The second makes use of accumulating regularizers centered at different approximate solutions, which we sequentially minimize in order to provide our near-optimal rates.
Read the full abstract
Networking
Sentinel-Based Failover for QKD-Augmented IPsec Tunnels
Quantum-safe IPsec through hybrid key establishment is practical, but creates a critical operational challenge: how to maintain tunnel availability when the QKD infrastructure becomes unavailable. In this paper, we present the design, implementation, and experimental evaluation of a quantum-safe key establishment mechanism for an IPsec tunnel that combines X25519, ML-KEM, and ETSI GS QKD 014 keys through the RFC 9370 multiple key exchange mechanism, and that degrades gracefully when the QKD key delivery fails. Our open-source StrongSwan plugin uses a sentinel-based coordination protocol, thereby permitting us to complete the handshake even if the QKD leg fails, instead of aborting, restoring the QKD share at the next rekey. On a testbed connected to a metropolitan QKD link over 33 km of deployed fiber, we evaluated five configurations, from a classical X25519 with RSA baseline to a hybrid one that adds ML-KEM-1024 and a QKD key. The full hybrid authentication costs 103 ms against 61 ms for the baseline, the QKD retrieval itself adds only about 7 ms. Failure injection experiments confirm that the tunnel survives a complete KME outage without any interruption of the protected traffic.
Read the full abstract
Robotics
DNC-IMM: Early Lane-Change Intention Recognition via Neural Calibration Based on Driving Context Information
Early recognition of lane-change intention is essential for proactive decision-making in autonomous driving and advanced driver assistance systems. This paper proposes a Dual Neural-Calibrated Interacting Multiple Model (DNC-IMM) that improves adaptability to driving context while preserving the probabilistic structure and interpretability of a conventional IMM. The proposed method encodes driving-context information, including target-vehicle motion, gaps to surrounding vehicles, and relative velocities, with a neural network that calibrates both the transition-probability matrix and measurement likelihoods. The final intention is determined from the calibrated IMM mode posterior rather than from a separate direct classifier. Experiments on the highD dataset demonstrate that the proposed method reliably recognizes lane-change intentions before lane crossing and provides particularly strong performance at the earlier 2-3 s prediction horizons.
Read the full abstract
Solar and Stellar Astrophysics
Integrated Mass Loss for Very Metal-poor Stars: Asterosesimic Red Giant Masses of K2 Globular Cluster NGC 5897
Mass loss in low-mass stars during the red giant branch (RGB) and early asymptotic giant branch (EAGB) phases plays a key role in shaping stellar evolution, yet its dependence on stellar parameters such as metallicity remains poorly constrained, with observational studies yielding conflicting trends. We present the first asteroseismic analysis of RGB and EAGB stars in NGC 5897, the most distant and metal-poor globular cluster observed by the Kepler space telescope during the K2 mission. We detected solar-like oscillations and derived the frequency of maximum power excess, $\nu_{\rm max}$, for 20 RGB and 6 EAGB stars. Using asteroseismic scaling relations, we derived mean masses of $\overline{M}_{\rm RGB} = 0.74\pm0.01\,M_{\odot}$ and $\overline{M}_{\rm EAGB}=0.65\pm 0.03\,M_{\odot}$. The inferred integrated mass loss between the two phases is $\Delta M_{\rm RGB-EAGB}=0.08\pm 0.03 \,M_{\odot}$. We present an updated mass-loss–metallicity relation for Type I globular clusters, extending it to the very metal-poor regime and supporting decreasing integrated RGB mass loss with decreasing metallicity.
Read the full abstract
Mesoscale and Nanoscale Physics
Interfacial orbital torques excite nanoscale terahertz magnons
Exchange-dominated magnons in nanometer-thick ferromagnets extend to the terahertz regime through thickness quantization of perpendicular standing spin-wave (PSSW) modes. While interfacial spin-orbit torques (SOTs) have been shown to enable the excitation of such modes, the microscopic origin of the interfacial driving torque remains unclear. In particular, the coexistence of spin and orbital currents complicates the understanding. Here, we develop and use an atomistic framework that explicitly resolves interfacial symmetries and separates spin and orbital torque contributions. Exploiting a trilayer geometry for a thin ferromagnet sandwiched between non-magnetic layers, where the symmetry-controlled polarity of the interfacial torque produces mode-selective magnon excitation as observed in the recent experiment of Salikhov et al. Nature Phys. 19, 529 (2023), we disentangle the different interfacial torque contributions. By decomposing the torque into magnetization-even (field-like) and magnetization-odd components, we identify the field-like torque as the dominant contribution responsible for the excitation. Crucially, isolating orbital and spin contributions reveals that the interfacial orbital torque provides the primary channel for the efficient excitation of exchange-dominated THz magnons in thin ferromagnets. Our results establish a microscopic basis for symmetry-engineered control of confined terahertz spin dynamics in magnetic multilayers.
Read the full abstract
Artificial Intelligence
Latent Recurrent Thoughts: Recurrent Refinement of Proposed Latents for Reasoning with Frozen LLMs
Chain-of-thought reasoning unfolds in discrete token space: each step is committed as text, errors propagate, and eliciting good traces presupposes traces to imitate. Reasoning instead in a model's continuous representation space - where intermediate states are vectors rather than words - sidesteps these constraints, but leaves open how those latent states should be computed. We approach this along two axes. First, we keep a large language model (LLM) frozen and use it for what it is already good at - modeling and decoding sequences - while a small auxiliary network supplies continuous latent thoughts as input. Second, we produce those latents by recurrence: a tiny recurrent reasoner refines them over many steps, decoupling the depth of computation from the size of the model, so that the latents are a product of iterative processing rather than a single forward pass. We instantiate this as Latent Recurrent Thoughts (LRT): a task-dedicated proposer supplies base latents, a recurrent reasoner refines them through bounded residual corrections, and the frozen LLM decodes the answer. On symbolic reasoning with answer supervision but no reasoning traces (Countdown-4, Sudoku) and on natural-language reasoning (HumanEval, MBPP, StrategyQA), LRT substantially outperforms prior frozen-decoder continuous-space reasoning methods under an identical decoder, prompt, data, and training budget, and outperforms non-thinking-mode chain-of-thought prompting on the same backbone at a small fraction of its inference compute.
Read the full abstract
Functional Analysis
Vector-Carleson, Calderón, and Poisson-Atomic Criteria for Generalized Hilbert Operators on $H^p$, $p>2$
Let $\mathcal H_g f(z)=\int_0^1 f(t)g'(tz),dt$, and let $2<p<\infty$. Set $t=2p/(p-2)$ and $X_j=2^{-j/p'}\Delta_j g'$, where $\Delta_j$ is the hard dyadic Taylor projection. We prove that $\mathcal H_g:H^p\to H^p$ is bounded if and only if $h\mapsto(X_jh)_{j\geq0}$ is bounded from $H^t$ to $\ell^t(H^2)$. The square of this embedding norm equals the norm of the positive column operator $b\mapsto\sum_j b_j|X_j|^2$ from $\ell^{p/2}$ to $L^{p/2}$. Coordinate tails yield essential-norm estimates and an exact compactness criterion, while Hardy duality gives an equivalent paraproduct formulation. The criterion is quantitatively invariant under admissible analytic dyadic resolutions and defines a resolution-independent Calderón symbol space equal to the Hilbert-range multiplier space. We construct a bounded noncompact dense-frequency symbol outside the known blockwise sufficient class. We also prove an exact Poisson-atomic testing theorem: finite positive Poisson mixtures recover the full norm, but no fixed atom count suffices. Finally, aggregate probability densities give an intrinsic atomic-complexity formula and a finite-bandwidth testing bound.
Read the full abstract
Signal Processing
32-point DFT Approximations Based on Minimal Frobenius Error and DFT Symmetries
This work introduces low-complexity, multiplierless approximations for the 32-point discrete Fourier transform. The proposed methods are obtained by minimizing the Frobenius error compared against the DFT matrix over a set of trivial multipliers. A row-wise, symmetry-constrained parameterization is employed to reduce the search space size, rendering the task computationally tractable. The resulting approximations could outperform the reference method in the literature according to energy-based error measurements. A sparse matrix factorization is provided for efficient computation; the arithmetic costs are 152 real additions and 34 bit-shifts only.
Read the full abstract
Combinatorics
Unbalanced spectral Turán problem for color-critical graphs with prescribed large maximum degree
Let $F$ be a connected color-critical graph with $\chi(F)=r+1\ge4$, let $S_{n,\Delta}^{(r)}=(n-\Delta)K_1\vee T(\Delta,r-1)$. We determine the graph of maximum adjacency spectral radius among all $n$-vertex $F$-free graphs with prescribed maximum degree $\Delta$. There is a constant $s_F\in[0,1)$ such that, for all sufficiently large $n$, $\left\lceil\frac{(r-1)n}{r}\right\rceil\le \Delta\le n-\Theta(n^{s_F})$ implies that every $n$-vertex $F$-free graph $G$ with $\Delta(G)=\Delta$ satisfies $\rho(G)\le \rho\bigl(S_{n,\Delta}^{(r)}\bigr)$, with equality if and only if $G\cong S_{n,\Delta}^{(r)}$. This is the spectral counterpart of the edge theorem of [European J. Combin. 106 (2022), 103576.] and extends the clique result in [arXiv:2608.26634, 2026.]. This result also provides a benchmark for unbalanced spectral Turán problems arising from other extremal parameters.
Read the full abstract
Computation and Language
EDRAC: Benchmarking Arabic Dialect Reading Comprehension
Dialectal Arabic (DA) remains under-resourced compared to Modern Standard Arabic (MSA), particularly for machine reading comprehension (MRC) and question answering (QA). Existing Arabic QA benchmarks primarily focus on formal written MSA or multiple-choice QA, with limited coverage of naturally spoken dialects. Here, we aim to bridge this gap. We introduce EDRAC, the first large-scale benchmark for dialectal Arabic machine reading comprehension (MRC) and generative QA, covering five major dialects: Egyptian, Moroccan, Emirati, Syrian, and Saudi Arabic. EDRAC contains 499 passages derived from naturally occurring spoken interactions and 4,977 corresponding QA pairs generated through a human–LLM collaborative pipeline combining iterative generation, LLM-as-a-judge evaluation, and human verification. We benchmark Arabic-centric and multilingual LLMs on EDRAC using lexical and semantic metrics. Our results reveal substantial gaps between semantic answer quality and dialectal fidelity, highlighting the limitations of existing evaluation metrics for dialectal Arabic generation. EDRAC provides a realistic and challenging MRC benchmark for future research on dialectal Arabic NLP.
Read the full abstract
Information Theory
Generalized Tan-Arlery-Rabaste-Lehmann-Ovarlez Lower Bound on Ambiguity Function of a Set of Sequences With Mismatched Filters
In this paper, a lower bound on the maximum ambiguity function (AF) sidelobes of a set of unimodular sequences is formulated for the desired low-ambiguity-zone (LAZ). Our main idea is to introduce a set of mismatched filters associated to a set of unimodular sequences and two weight vectors for the delay and Doppler shifts, respectively. The length of mismatched filter maybe different to the length of unimodular sequence. The proposed lower bound on the maximum AF sidelobes for the desired LAZ can be treated as an extension of Tan-Arlery-Rabaste-Lehmann-Ovarlez lower bound, published in 2020, which dealt with the conventional correlation of sequences.
Read the full abstract
Computation and Language
ClinTraceBench: Source-Verifiable Longitudinal Clinical Reasoning over EHR-Derived Dialogues
Clinical LLM assistants must reason over multi-visit patient trajectories, yet whether the compact history representations used to scale them—retrieval, structured timelines, LLM summaries, agentic memory—preserve the longitudinal signal clinical reasoning needs has not been measured. We introduce ClinTraceBench: 385 MIMIC-IV-derived verified dialogues with event-ID provenance, a nine-task taxonomy (T1–T9), and L0–L4 deterministic + L5 human-audit validation (98.92% agreement). We evaluate eight history representation strategies—a no-context floor, last-visit-only, full-context, BGE-M3 dense-retrieval, two compression schemes, and two agentic-memory systems (Mem0, A-Mem)—across four backbones (DeepSeek-V3, GPT-4o-mini, Haiku~4.5, Sonnet~4.6) on 6{,}271 questions: 32 cells, 200{,}672 predictions. Four findings: (SP4) a controlled T3 injection probe isolates compression-induced relation loss—with the attribution sentence present before construction, Mem0, A-Mem and llm-summary still recover only 0–5.3% of the injected positives; (SP1) compressed strategies pay an aggregation tax on multi-visit trends and cross-patient comparisons; (SP2) the blind-to-full gap spans $+29.8$~pp (GPT-4o-mini) to $+62.7$~pp (Haiku); (SP3) abstention scales non-monotonically with context length. On the Pareto frontier Haiku dominates Sonnet under full-context (\$25.76 vs.\ \$106.21), inverting the “biggest backbone wins” heuristic.
Read the full abstract
Programming Languages
Griotte: Verified Compartmentalisation via Capabilities
CHERIoT is a novel hardware-software co-design that leverages hardware capabilities to define a notion of compartment, in a minimalistic capability-based OS, CHERIoT RTOS. By default, compartments are isolated to limit damage in case of bugs or malicious behaviour. To allow cross-compartment communication, the OS provides a privileged component, called the switcher. The switcher provides an interface for cross-compartment calls, while enforcing isolation between compartments and guaranteeing stack safety. Together with hardware capabilities, the switcher is critical to enforce the security guarantees of the CHERIoT compartment model. The design of CHERIoT raises two questions: First, how can one formalise the informal notion of compartmentalisation that CHERIoT compartments are designed to provide? And second, given that the safety properties of CHERIoT hinge on the complementary roles of the capability machine and of the switcher, does the design of CHERIoT enforce the desired security properties? In this paper, we introduce Griotte and Griotte OS, idealised but faithful versions of the CHERIoT machine and the CHERIoT RTOS, which we use to answer these two questions: First, we formally capture the aforementioned security guarantees in the form of a continuation-based logical relation which captures the combined behaviour of the switcher and of the capability machine. And second, we define a specification for the Griotte switcher that enforces those guarantees, and prove that the implementation meets the specification. We demonstrate Griotte on a range of key scenarios illustrating different aspects of CHERIoT, including integrity of the local state in the presence of memory sharing with unknown code. Our approach is modular: we verify compartments individually, and then compose their specifications. Together, our contributions give a solid formal foundation to the design of CHERIoT.
Read the full abstract
High Energy Astrophysical Phenomena
Complex Lags from Simple Physics
X-ray timing information produced through Fourier analysis from the variable emission of black hole X-ray binaries has been used for several decades to provide key insights into the physical setup of these systems not measurable with spectroscopy. In particular, quasi-periodic oscillations within these systems have been of particular interest and remain the source of great debate on how they come about. We investigate the timing products of simple toy models of QPO variability to provide more intuition when thinking about signals produced by these sources. We simulate simple physical setups and show how phase lags and coherence of the signals change in these different setups. We first focus on properties of QPO like signals under a single driving signal assumption. We then investigate the case of multiple oscillations in a signal. Finally, we investigate how timing products change when QPOs are produced by time dependent modulation of periodic signals. Many simple physical setups with common driving signals are able to reproduce complex non-linear phase lags that resemble those present in the data. The changes in physical setup aligns with experience in the data, such as differing power spectra but fall short of reproducing the data as expected. Multiple incoherent processes present in the signal struggle to reproduce behaviour present in the data. Coherence seems to be a more useful tool for differentiating between setups. Fourier analysis with complicated data like that produced by X-ray binaries can lead one to be tempted to invoke exotic lag mechanisms without the appropriate framing. This paper attempts to help provide tools and intuition as to how different phenomena in signals (particularly relating to QPOs) can result in non-linear phase lags with explicit structure.
Read the full abstract
Machine Learning
Replicating TRACE: A Practitioner's Guide to Its Threshold and Particle Budget
TRACE (Math & Lienhart, arXiv:2602.01135) reads causal graphs over event types out of a pretrained autoregressive sequence model by thresholding a per-position conditional-mutual-information estimate at a fixed tau. We independently replicate its headline synthetic result: with tau selected on a validation split, mean per-sequence F1 against exact interventional truth reaches 0.90-0.91 at vocabulary size 1000 (paper: 0.91) and 0.86-0.91 from 100 to 2000. First, the optimal threshold is pinned to the truth margin, not to any constant: at every size the errors at tau* straddle the delta = 0.05 margin defining ground truth (missed true edges lie just above it, accepted false ones just below), and the blind optimum lands near delta/2 times the estimator's calibration, confirmed out of sample at 5000. Second, at a single global threshold TRACE mostly recovers a direct, adjacent-influence graph: lag-1 true edges are recalled at 0.97-0.99, while true edges at lag 2 or more read orders of magnitude lower—the reading-scale price of randomizing mediating positions, which an exact test of direct causal effect requires when the truth is unknown. A per-lag threshold family recovers a third to a half of lag-2 truth; on lag-uniform data one validated threshold recalls every lag at 0.40-0.87, 8-26 pp below an atomic-intervention control at lags 3-6. Third, the default lag decay of the paper's synthetic benchmark concentrates about 85% of interventional truth at lag 1 and pushes the rest below the estimator's noise floor, so headline F1 there certifies lag-1 recovery only and conflates the benchmark's skew with the algorithm's own limit; a flatter decay separates the two. Fourth, F1 saturates from N = 2 particles at the selected threshold—a property of the threshold's margin over the noise floor, not of the estimator, which converges as N^(-1/2). We distill five practitioner rules.
Read the full abstract
Numerical Analysis
Symmetry Inheritance and Symmetry-Reduced Finite Element Analysis for Second-Order Linear Elliptic Problems with Robin Boundary Conditions
A symmetry framework is developed for second-order linear elliptic equations subject to Robin boundary conditions on bounded domains. Orthogonal transformations of the domain are represented through left group actions on scalar, vector, and second-order tensor fields. The corresponding transformation rules for the gradient, divergence, diffusion flux, and conormal boundary term are derived in detail. Based on these relations, symmetry groups are introduced for the principal coefficient tensor, the first-order and zeroth-order coefficients, the differential operator, the Robin coefficient, and the boundary operator. The symmetry properties of the volume and boundary source terms are then incorporated into a common symmetry group for the complete boundary value problem. Under the assumption of unique solvability, it is proved that every element of this common group is also a symmetry of the solution. For reflection symmetries, homogeneous generalized Neumann conditions are obtained on artificial symmetry boundaries, leading to an exact reduction of the computational domain. A corresponding finite element formulation is presented, and numerical results are provided to verify the theoretical symmetry properties and the validity of the resulting domain-reduction strategy. Three examples illustrate radial reduction from multiple dimensions to one dimension, reflection-based domain reduction, and a variable-coefficient problem in which all coefficients and source terms are nonzero.
Read the full abstract
Software Engineering
Hints Help But Do They Teach? Evaluating Skills Transfer in Code Generation
When a hint turns a failing generated program into a passing one, does it provide missing information or merely steer the model toward a solution it could already produce? We test these hypotheses on HumanEval+ and MBPP+ using executable evaluation. For Qwen2.5-3B-Instruct, adaptive relevant hints rescue 36 of 79 selected failures; an unrelated hint rescues 19, while eight unhinted samples solve 46 and recover 31 of the 36 relevant-hint rescues. Phi-3.5-mini shows the same pattern: relevant hints rescue 42 of 101 failures, an unrelated hint rescues 17, and unhinted sampling solves 57, including 36 of the 42 relevant-hint rescues. Because the hint conditions use different attempt budgets, these comparisons do not isolate a purely semantic effect. Mechanistic tests on Qwen identify a stable activation direction shared by relevant and unrelated hints. Persistently adding this direction yields 14 rescues and 18 regressions, with no detectable net accuracy gain; learned low-rank interventions have a positive but imprecise estimated effect. Full textual specifications solve 22 of 24 context-defined problems, versus 5-11 for tested virtual-KV prefixes. Post-generation hidden-state probes transfer across benchmarks, with pooled AUROC 0.806 and 0.780, but their top-one selection advantage over token confidence is statistically unresolved. Overall, relevant hints can rescue failures, but most rescued solutions are already reachable through ordinary sampling, and the internal interventions tested here do not establish task-general capability transfer.
Read the full abstract
Analysis of PDEs
Obstruction to quasi-invariance of Gaussian measures under transport flows on Riemannian Manifolds
We prove an obstruction to quasi-invariance of Gaussian fields under divergence free transport flows on Riemannian manifolds. For the centered Gaussian field with covariance $(1-\Delta_g)^{-\alpha}$, $\alpha>\frac{d}{2}$, quasi-invariance under the transport flow is equivalent to invariance. This happens precisely when the flow acts by isometries, or equivalently when the underlying vector field is Killing. The proof leverages the Feldman–Hájek theorem and utilizes localized pseudo-differential computations. On $\mathbb R^d$ this leaves only rigid motions, while on flat tori this leaves only translations.
Read the full abstract
Differential Geometry
The asymptotic Plateau problem for Hypersurfaces of constant $H_{k}$ curvature in hyperbolic space
In this paper, we study the asymptotic Plateau problem in hyperbolic space for hypersurfaces of constant $H_k$-curvature. We prove the existence of a smooth complete $k$-convex hypersurface in $\mathbb{H}^{n+1}$ satisfying \[ H_k(\kappa)=\sigma, \qquad \sigma\in(0,1), \] with prescribed asymptotic boundary at infinity. In particular, our result extends the range of the constant $\sigma$ in the existence theorem of Guan and Spruck [J. Eur. Math. Soc. 12 (2010), no. 3, 797–817] for $H_{k}$ curvature to the full interval $(0,1)$.
Read the full abstract
Computation and Language
When Modality Gap Reduction Fails: Prediction-Level Hubness in CLIP
Reducing the modality gap between image and text representations in CLIP is widely expected to improve cross-modal alignment and downstream performance. However, a smaller average image-text gap does not necessarily lead to consistent accuracy gains. We analyze this mismatch from the perspective of the decision structure in zero-shot classification, i.e. selecting the most similar class-text prototype for an input image. Zero-shot accuracy depends not only on average image–text alignment, but also on class-wise decision margins. Using Linear correction as an analytically tractable case, we show that modality gap correction can alter the relative decision structure among classes and cause predictions to concentrate on a small subset of classes. We refer to this output-space failure mode as prediction-level hubness. Furthermore, experiments across multiple datasets show that accuracy degradation under gap correction is consistently associated with increased prediction concentration, both for Linear correction and for learning-based correction methods. This provides a systematic explanation of why modality gap reduction does not consistently improve CLIP zero-shot accuracy from the perspective of downstream decision structure. Our results suggest that gap correction should be evaluated not only by average alignment, but also by its impact on downstream prediction structure.
Read the full abstract
Machine Learning
Neural Symbollic Regression Using Deep Learning and Sparse Modelling
Symbolic Regression (SR) seeks to find succinct mathematical expressions that represent the fundamental relationships within data, providing interpretability and scientific understanding that exceeds that of black-box models. Nevertheless, traditional methods like Genetic Programming face challenges with scalability and are highly sensitive to noise, while sparse regression techniques such as SINDy rely significantly on predetermined feature libraries. In this work, we present a Neural Symbolic Regression (NSR) framework that treats neural networks as functional preconditioners for symbolic discovery. Our approach uses a decoupled pipeline: a neural network first learns a smooth, noise-robust approximation of the target function in an interaction- aware nonlinear feature space. LASSO is then applied to extract sparse, interpretable closed-form expressions. To improve predictive accuracy and symbolic fidelity by integrating distributed hyperparameter optimization with Ray Tune and ASHA scheduling. Experiments on the Nguyen benchmark suite show that our approach consistently outperforms SINDy and non-tuned neural baselines in RMSE, noise robustness, and out-of-distribution generalization. Ablation studies confirm the significance of feature interactions, neural depth, and tuning strategies. In general, this study presents a scalable and understandable neural-symbolic framework, creating a solid link between neural approximation and the discovery of sparse equations for scientific machine learning.