High Energy Physics (Phenomenology)
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Tuesday 10 November
Motivated by the experimental observations of $f_4(2050)$ and $f_4(2300)$ and the establishment of $a_4$ family, we investigate the $f_4$ family by combining the mass spectrum under the modified Godfrey–Isgur model with two-body strong-decay calculations in the quark-pair-creation model. The common pair-creation strength is fixed by the measured total width of $f_4(2050)$. The mass spectrum analysis and decay information both favor the nonstrange $2\,{}^3F_4$ assignment for $f_4(2300)$. The largely strange $2F$ partner is predicted at $2453$ MeV with a total width of $216$ MeV, decaying mainly into $K^*\bar K^*(1410)$ and $K^*\bar K^*$. Predictions are also given for the $3F$, $4F$, $1H$, and $2H$ flavor pairs. At the adopted mixing angles, the $H$-wave states are more than twice as broad as nearby $F$-wave states of the same flavor. The results of this work may serve as useful guidance for future experimental studies.
Systems and Control
Aeroelastically Constrained Optimal Path Planning of a Fixed-Wing Airborne Wind Energy System
Multi-megawatt fixed-wing airborne wind energy systems (AWES) use an aircraft with large, slender wings that deform over the pumping cycle. To harvest maximum energy, the cycle is usually designed by means of dedicated reference trajectory optimization routines. It is well known that such routines rely heavily on the system model and that close alignment between the model and the physical system is required for the results to hold in practice. Nevertheless, the wing is usually treated as rigid and structural limits are respected indirectly through upper bounding of states and inputs. In this work, we directly enforce bounds on the elastic deflection and twist of the wing in the power-maximizing periodic optimal control problem. To that end, we model the wing by root-clamped modes, loaded by strip-theory aerodynamics and by the gravitational and inertial loads of the manoeuvring aircraft. To keep the problem tractable, the modal dynamics are replaced by their quasi-static equilibrium, which enters every discretization interval as algebraic constraints. For a multi-megawatt reference aircraft, the elastic twist near the wing tips reaches $1.03^\circ$ along the optimal rigid-wing trajectory. With a twist limit of $0.6^\circ$, the optimal trajectory delivers $31\,\%$ less mean power than the rigid-wing one, mainly through a lower tether tension during reel-out. Rigid-wing trajectory optimization can thus overestimate the power that the wing limits permit. Forward simulations that retain the modal dynamics demonstrate that the quasi-static deflection and twist deviate from the dynamic ones by less than $2\,\%$ of their limits. This work shows that elastic deformation can be accounted for in the trajectory optimization and that, depending on the wing limits, it can strongly reduce the achievable power.
Read the full abstract
Graphics
Deformable Texture Maps: Compiling Dynamic Microgeometry into Neural Textures
Deformable material surfaces, such as cloth, owe their appearance to fine-scale geometries that restructure under deformation. Even when coarse simulation models efficiently recover this microstructure, explicitly instantiating and rendering it remains expensive. We introduce Deformable Texture Maps (DTM), a representation that compiles deformation-driven microgeometry into a continuous and compact neural texture. Our model maps texture coordinates and local deformations to intrinsic maps describing opacity, surface normals, depth, and material identity. This captures the occlusion and visibility changes generated by deformation. At runtime, these maps provide a common texture-space representation of deformation-driven fine-scale structure that can be evaluated directly on coarse deforming surfaces. Since our representation encodes geometric attributes rather than radiance, lighting and material response remain decoupled from the learned model. This allows us to apply relighting, as well as shading and material editing without retraining. We further introduce a persistent texture-space rendering strategy that caches these intrinsic maps and selectively updates them as deformation changes. This effectively decouples neural evaluation from frame rendering and enables their reuse across frames and conventional rendering effects. We demonstrate our approach on yarn-based textiles, where non-linear material stretch and shear induce changes in yarn geometry, spacing and occlusion. We build our model by, first, sweeping tileable yarn patches through deformation space, then, rendering the detailed geometry into a canonical texture-space, and finally, compressing the resulting maps into a single neural representation. Our resulting model can be evaluated directly on arbitrary coarse garment meshes and integrated with conventional shading pipelines, keeping appearance authorable at render time.
Read the full abstract
Combinatorics
An automated approach to structure theorems for minimally repetitive rich words
Recent work on repetitions in rich words has led to the intensive study of the family of infinite rich words $\{{\mathbf u}_d:d\geq 2\}$, where ${\mathbf u}_d$ is the fixed point of the morphism $\varphi_d:\Sigma_d^*\rightarrow \Sigma_d^*$ defined by $\varphi_d(i)=\tt{0}(i+1)$ for $i\in\{\tt{0},\ldots,d-2\}$ and $\varphi_d(d-1)=\tt{0}(d-1)(d-1)$. For $d=2$ and $d=3$, it is known that the repetition threshold for infinite $d$-ary rich words is equal to the asymptotic critical exponent $E^*({\mathbf u}_{2d-1})$. Moreover, the structure of ${\mathbf u}_{2d-1}$ underlies the structure of all “minimally repetitive” $d$-ary rich words: there is a real number $t_d>E^*({\mathbf u}_{2d-1})$ such that every infinite $t_d$-power-free $d$-ary rich word contains the language of $\pi({\mathbf u}_{2d-1})$ for one of finitely many morphisms $\pi:\Sigma_{2d-1}^*\rightarrow \Sigma_d^*$. We conjecture that a similar situation holds for all $d\geq 2$, and prove this for the first unknown case $d=4$. To do this, we develop a new approach for proving such structure theorems. Two key new concepts are “compatibility” of return words to palindromes in recurrent rich words, and “$t$-persistent repetitions”. They facilitate the automation of several aspects of the proofs which were previously done laboriously by hand. We first demonstrate our approach by giving a much shorter proof of the known structure theorem for infinite $\tfrac{16}{7}$-power-free ternary rich words. Then we prove a new structure theorem: every infinite $2.12$-power-free quaternary rich word contains the language of one of seven distinct words of the form $\pi({\mathbf u}_7)$, up to permutation of the letters. We prove that all seven words are rich and have critical exponent $E^*({\mathbf u}_7)$, so the repetition threshold for quaternary rich words is $E^*({\mathbf u}_7)\doteq 2.1197196907$.
Read the full abstract
Computation and Language
OpticalRAG: Pixel-Space Compression for Token-Efficient Retrieval-Augmented Generation
Retrieval-augmented generation (RAG) is bottlenecked by the token budget of large language models: feeding long documents into the context window is expensive, while aggressive textual compression discards fine-grained information that is irreversibly lost in a one-dimensional token sequence. We argue that this bottleneck is not fundamental to information density but to the choice of compression modality. Modern visual encoders can pack the contents of a rendered text page into roughly a hundred visual tokens with little semantic loss, opening a new compression-fidelity tradeoff that purely textual methods cannot reach. We present OpticalRAG, a framework that renders documents as page images, encodes them with a frozen vision encoder, and distills the retrieved visual tokens into as few as $16$ injected tokens before passing them to the LLM. Two ingredients make this work without expensive cross-modal training: (i) Encoder-Bridged Retrieval, which projects text-query states into the visual feature space of the frozen encoder, through a projection shared with the compressor, to retrieve visual tokens from a text query, with no contrastive training; and (ii) Query-Driven Distillation, a lightweight cross-attention compressor that condenses the retrieved visual tokens into a small set of query-conditioned prefix tokens. On LongBench-v2, BAMBOO, and LooGLE-v2, OpticalRAG achieves the best compression-aware accuracy (CAP) on all three benchmarks while injecting only $16$ tokens, and improves the closed-book Qwen2.5-7B-Instruct backbone by $5.17$ points on LongBench-v2. Code is available at https://github.com/brighterr/OpticalRAG.
Read the full abstract
Analysis of PDEs
Uniqueness and Nondegeneracy of Ground States for Critical-Hardy Fractional Equations
We study ground states of nonlinear fractional Schrödinger equations with the critical Hardy potential in $\mathbb{R}^d$, where $d\geq3$. We prove the existence and qualitative properties of ground states, based on which we analyze the uniqueness and nondegeneracy of ground states. As a byproduct, we also investigate the optimizers of the critical Hardy–Gagliardo–Nirenberg inequality, from which a sharp Keller–Hardy–Lieb–Thirring inequality is analytically derived.
Read the full abstract
Systems and Control
Priority-Aware Firm and Flexible Capacity Quantification for Large-Load Interconnection
We study how to unlock existing transmission capacity to accelerate large-load interconnection under differentiated quality-of-service (QoS) requirements. We establish an efficient framework to quantify firm and flexible hosting capacities and assess marginal values of grid upgrades. Specifically, we define multiple flexible capacity tiers with differentiated QoS limits and design a two-layer priority rule to allocate curtailment among tiers. We then formulate a mixed-integer chance-constrained model to jointly quantify firm and flexible capacities under QoS and priority constraints. To improve computational efficiency, we derive an equivalent aggregate formulation that eliminates individual curtailment variables and priority binaries without changing the feasible capacity set. Numerical results show the computational advantage of aggregation. Multi-tier flexible capacity increases total admitted capacity by 45.0% over the firm-only hosting capacity.
Read the full abstract
Systems and Control
Efficient control of multi-agent robust MDPs through tensor trees
This paper introduces a tensor-based framework for controller synthesis of stochastic multi-agent systems under temporal logic specifications that couple the agents, such as one agent reaching its goal only after another agent has reached its own. The agents are modeled as robust Markov decision processes (RMDPs), whose transition probabilities are only known to lie in an ambiguity set, for instance due to abstraction errors, learned models, or uncertain parameters. We show that the decoupled nature of the ambiguity sets allows one to decompose the computation of a lower bound on the satisfaction probabilities into per-agent factors, so that only the per-agent ambiguity sets need to be stored. This leads to improved memory complexity compared to monolithic approaches. We compare our approach with state-of-the-art approaches based on approximate simulation relations and with interval MDP abstractions. On a four-agent ordered reach-avoid benchmark, our approach certifies a lower bound twice as high as the simulation relation- based approach with 20 times less memory, and needs about 200 times less memory than the interval MDP abstraction, even at a resolution with about a million times more joint cells.
Read the full abstract
Quantum Physics
Quantum-Assisted Generative Anomaly Detection: A Dimension-Controlled Study of Latent Refinement
Anomaly-based malware detectors learn what benign files look like and flag deviations. Generative approaches to this problem can make use of a "latent" or "prior" data representation, which is often chosen to be a Gaussian distribution. We ask whether the choice of prior distribution matters, and whether one produced by quantum hardware behaves differently from a classical one. Building on SAGEGAN, a style-based detector trained only on benign data, we compare three latent distributions: the original 512-dimensional Gaussian, a 48-dimensional Gaussian, and a 48-dimensional distribution sampled from an ORCA Computing PT-2 photonic quantum processor. We then refine each trained model's frozen representation using greedy and reinforcement-learning methods. Across ten random seeds, the quantum-derived model shows richer internal structure with better effective rank and latent similarity. It also benefits most from refinement. Though final detection accuracy is similar to that of the refined Gaussian latent, these findings show differences in latent geometry and refinement behavior between the trained configurations.
Read the full abstract
High Energy Physics (Theory)
Geometry of Gauge-invariant Quantum Fields
Gauge-invariant representatives of massless gauge fields generally involve inverse differential operators. We trace this non-locality to the homotopy contracting the free gauge complex and replace it, at field level, by a local map using an independent scalar and a differential constraint. We discuss the BRST completion needed to preserve the physical content and the relation to the Stückelberg theory. We finally comment on the extension of this approach to non-Abelian gauge groups via the extension to interactions through $L_\infty$ homotopy transfer.
Read the full abstract
Artificial Intelligence
A computational framework for understanding embodied plans
Reasoning about people's plans and goals based on their actions is fundamental to human social interaction and essential for cooperative, embodied artificial intelligence. Formalizing this capacity in naturalistic, embodied settings has been a longstanding challenge. Here, we present GenEA, a novel agent representation capable of reasoning over multi-step, long-horizon embodied plans. At the heart of GenEA lies a multilevel state-space that couples the coarse biomechanics of an articulated body and its interactions with nearby objects by extending the task-and-motion planning framework from robotics with simulation-based intuitive physics. By embedding this state-space in a generative model, GenEA yields a joint probability distribution over goals and embodied plans. Across several experiments spanning the qualitatively distinct domains of object manipulation and navigation, featuring diverse tasks and multiple behavioral measures, we show that GenEA, through simple transformations of this joint distribution, recapitulates how humans reason about embodied plans. Crucially, GenEA consistently outperforms strong alternative models, including plausible heuristics, standard task-and-motion planning, and a state-of-the-art vision-language model. GenEA thereby offers a new perspective on the architecture of human reasoning and social cognition, suggesting that embodied artificial intelligence systems seeking to seamlessly interact with humans should adopt GenEA-like generative models of action planning.
Read the full abstract
Optimization and Control
Proximal point algorithm for interval valued optimization problems with application to lasso problem
In this paper, we consider a class of unconstrained interval valued optimization problems and using a suitable scalarization technique, we formulate an associated scalarized optimization problem and establish the relationships between the optimal solution of the scalarized problem and the $LU$-nondominated solution of the considered optimization problem. Based on this framework, we propose a proximal point algorithm and derive the global convergence of the sequence generated by the proposed algorithm along with the sublinear and linear rate of convergence for convex and strongly convex cases, respectively. Finally, we provide suitable numerical experiments along with comparative analysis with an existing method and applicability of the proposed algorithm on an interval valued lasso regression problem, highlighting its potential in machine learning models with uncertain data.
Read the full abstract
Artificial Intelligence
Benchmarking LLM Next-Step Recommendations with Knowledge Space Theory
Selecting an LLM for next-step recommendation requires benchmarks that reflect the intended educational task and domain. We present a benchmarking method that uses Knowledge Space Theory (KST) as a formal evaluation reference. Given a task catalogue and an existing prerequisite relation, the method derives admissible knowledge states and the complete set of immediate next steps for each state. Recommendations generated by a model under a specified prompt are compared with these reference sets. A sample evaluation study uses 2,400 responses from six models across four mathematics domains, four prompt conditions, and two catalogue orders. Exact-set accuracy was 61.3% with descriptions only, 62.8% when models were asked to reconstruct prerequisites, 96.3% with supplied prerequisite relations, and 98.2% with the complete knowledge space. Performance varied across models and domains, and recommendation sets agreed across the two presentations in 85.25% of pairs. The benchmark supports comparisons of model-prompt combinations against a context-specific reference. Its scores describe structural recommendation accuracy under the evaluated conditions; they do not establish learning benefit or the necessity of using an LLM for selection.
Read the full abstract
Probability
Tail Asymptotics for Gaps in Unitary Ensembles and the Airy Point Process
We derive tail asymptotics for the gap between the $k$th and $(k+1)$st largest particles, with $k$ fixed, in unitary ensembles and in the Airy point process. For the finite-$N$ ensembles, we consider the probabilities of a rescaled gap larger than $a_N$ and of a rescaled gap smaller than $b_N$, in the joint limits $N\to\infty$, $a_N\to\infty$ with $a_N=o(N^{2/3})$, and $N\to\infty$, $b_N\downarrow0$, respectively. For the Airy point process, the right and left tails are analyzed as the gap size tends to infinity and zero, respectively. More precisely, in both models, we obtain logarithmic right-tail asymptotics up to the constant term and cubic left-tail asymptotics with coefficient $k/6$. As an application, we also derive the limiting conditional laws of the leading particles given the existence of an abnormally large or small gap. In particular, for $k=1$, large-gap conditioning leads to a transition from Tracy–Widom to Gaussian fluctuations of the largest particle, similar to the BBP transition. Further, among any fixed number of leading gaps, we also determine the limiting probabilities of which gap attains a large maximum or a small minimum.
Read the full abstract
Quantum Physics
Observation of strong electric dipole-dipole interactions between two atoms
We have prepared pairs of atoms at densities exceeding $ 10^{16}/\rm{cm}^3$ corresponding to $n\lambdabar^3>10$ and studied how they interact with light. In this regime, the coherent part of the electric dipole-dipole interaction is larger than the dissipative part, and a single photon emission can transfer large momenta (larger than ten times the single photon recoil) to the atoms. For off-resonant laser light, the atoms are dressed with an electric dipole moment of $\approx$ 1 Debye and interact coherently. This light-induced dipole-dipole interaction has been previously observed only in the far-field mean-field regime. For two atoms, the resonance is shifted by the two-body Lamb shift which is dominated by the near-field part of the dipolar interaction. Dipolar resonance shifts have been observed before only for many atoms interacting with the far-field part of the dipolar interaction. Our work sheds new light on photoassociation processes and paves the way for studying driven quantum systems with strong electric dipole-dipole interactions.
Read the full abstract
Metric Geometry
Convex monohedral tilings: a sharp neighbor bound and rigidity in the equality case
We study both side-to-side and non-side-to-side convex monohedral tilings of the Euclidean plane in which every tile has the same number of neighbors. Reflections are allowed, point contacts count as neighbors, and no periodicity assumption is imposed. Combining boundary estimates for sums of face curvatures, angle-incidence counts and side-length matching, we prove that the common neighbor number satisfies $N\le21$ and classify all equality cases. Equality holds precisely when the prototile is a $30^\circ$-$30^\circ$-$120^\circ$ triangle and the entire tiling is similar to the central subdivision of the standard equilateral triangular tiling. This resolves the open problem on the maximum common number of neighbors recorded in Adams's The Tiling Book.
Read the full abstract
Optics
Overcoming Dynamic Range Limitations in Computational Hyperspectral Imaging
Computational hyperspectral imaging offers compact spectral sensing, yet imaging high-dynamic-range (HDR) scenes remains challenging. Accurate spectral reconstruction requires encoded measurements to remain within the calibrated linear-response range and above the noise floor. Due to unequal spectral responsivities across the encoding mask, individual channels reach their reliability limits at vastly different incident signal levels. Measurement reliability therefore depends on both exposure time and encoding channel, making it challenging for a single exposure to preserve spectral information in highlights and shadows simultaneously. Here, we demonstrate a physics-guided multi-exposure fusion framework that operates directly in the encoded measurement domain. An analytical reliability prior and learned spatially varying, channel-specific adjustments jointly determine adaptive fusion weights. Measurement-domain spectral-angle supervision further constrains relative channel responses for downstream spectral inversion. The set-wise formulation accommodates varying numbers and configurations of input exposures. We further implement the framework on a 9-channel broadband-encoded stochastic hyperspectral camera, incorporating full-field sensor-response characterization and channel-aware reliability gating. Simulated HDR benchmarks show lower reconstruction errors than representative fusion methods, while controlled spectral measurements and practical HDR scenes confirm improved spectral fidelity and extended usable dynamic range. Measurements using an optical-density target further establish a 16.0-dB increase in spectral reconstruction dynamic range relative to the best fixed single exposure.
Read the full abstract
Robotics
NIF-VIO: Visual–Inertial Odometry Across Inertial and Non-Inertial Reference Frames
Conventional visual–inertial odometry (VIO) assumes that observations from cameras and an inertial measurement unit (IMU) are jointly consistent in an inertial reference frame. However, this assumption breaks down under non-inertial conditions. Inside accelerating or rotating carriers, visual observations of the interior reflect carrier-relative motion, while IMU measurements contain additional carrier-induced components, which can compromise conventional fusion. We therefore present NIF-VIO for localization relative to the local environment across inertial and non-inertial reference frames. It adapts visual–inertial fusion based on measurement consistency. As inertial conditions change, early constraint isolation is combined with local reoptimization to limit error propagation before inconsistency confirmation and repair the affected local geometry. Under non-inertial conditions, a compact model with temporally shared parameters is introduced to account for carrier motion while preserving inertial information that complements visual constraints. Experiments on public benchmarks and a self-collected dataset demonstrate that NIF-VIO achieves competitive performance under inertial conditions and improved localization accuracy and availability under non-inertial conditions.
Read the full abstract
Machine Learning
SepsisFlow: an event-stream foundation model for continuous risk warning and trajectory modelling in ICU sepsis
Sepsis is a leading cause of ICU death with heterogeneous onset and course, yet most warning models predict only whether sepsis will occur, leaving clinicians uncertain of its onset context and early trajectory once an alarm sounds. To answer these questions on one hourly clock from ICU admission, we built SepsisFlow: before onset, it combines a Mamba model pretrained on 2.57 million EHRs with about 1,700 hourly features in gradient-boosted trees to predict 6/12/24-h risk; at onset, it fuses Mamba and tabular encoders via a gated network into continuous patient states defining onset contexts; after onset, it encodes 72-h organ-support and physiological trajectories via an attention bidirectional recurrent network into early courses. For pre-onset warning in MIMIC-IV, SepsisFlow detected more cases than NEWS2, qSOFA, and SIRS at equal alarms, gaining about 8 cases per 100 from hourly updating, while outperforming 16 retrained algorithms and maintaining risk ranking in eICU. At onset, its contexts separated 2% cardiovascular-ICU from 15%-17% other 28-day mortality, and within each context after onset, persistent support carried fivefold higher mortality than de-escalation. SepsisFlow establishes a foundation-model paradigm for continuous risk and trajectory stratification in critical care.
Read the full abstract
Combinatorics
Judicious $r$-partitions of $r$-uniform hypergraphs
Bollobás and Thomason conjectured that every $r$-uniform hypergraph with $m$ edges admits an $r$-partition such that each class meets at least $rm/(2r-1)$ edges. In this paper, we establish a lower bound of $(11-\sqrt{13})m/18$ for every $r\ge 2$. This improves Haslegrave's bound $rm/(3r-4)$ for every $r\ge 8$.
Read the full abstract
Analysis of PDEs
Global existence and blow-up for a fractional wave equation with logarithmic nonlinearity and strong damping
This paper studies a fractional strongly damped wave equation with logarithmic nonlinearity. Based on local well-posedness of weak solutions via the Galerkin approximation, we employ the potential well theory to derive the existence of global solutions under subcritical and critical initial energy conditions. Moreover, infinite time blow-up phenomena are obtained under subcritical, critical, and arbitrarily high supercritical initial energy conditions.
Read the full abstract
Quantum Physics
An Efficient Quantum Algorithm for Preparing Uniform Superpositions over Arbitrary Integer Intervals
We consider exact preparation of the uniform quantum state supported on an arbitrary nonempty interval of integer labels in an $n$-qubit register. For every $0\leq a<a+M\leq2^n$, we give an ancilla-free circuit using only one-qubit gates and singly controlled one-qubit gates. In the logical gate model specified in the paper, the direct construction has exact worst-case gate count $5n-9$ for $n\geq3$; its exact maxima for $n=1$ and $n=2$ are $1$ and $3$, respectively. A worst-case lower bound of $\Omega(n)$ establishes asymptotic optimality in the register width. In addition, the gate complexity admits an instance-dependent bound $O(\log_2 M+\epsilon)$, where $\epsilon$ is an offset-dependent parameter defined in terms of binary Hamming weights. We also derive a reflected construction and a classical procedure that selects the lower-count exact circuit. The direct construction admits an $O(n)$ logical-depth upper bound. We also treat bitwise-XOR images of intervals and arithmetic progressions with power-of-two spacing.
Read the full abstract
Statistical Methodology
Projection-Based Partial Correlation Inference Beyond Sparsity
Valid inference for partial correlation in high-dimensional Gaussian models is challenging when the conditioning variables exhibit dense dependence and conventional sparsity assumptions are inappropriate. We develop a target-specific projection framework for estimating and testing a population partial correlation without requiring coordinate sparsity. A population quadratic criterion characterizes the information retained by a linear reduction and is maximized by a one-dimensional projection. Under a rank-one alignment condition, this optimal projection preserves the original partial correlation. To obtain a feasible procedure, we estimate the optimal direction using diagonal ridge regularization on an independent training sample. Under conditions on regularized effective dimension and target-specific ridge bias, we establish a quantitative direction-estimation rate in the covariance-induced prediction norm. We further show that, under alignment, the induced partial-correlation error is second order in the direction error. These results yield oracle equivalence and asymptotic normality for the feasible split-sample estimator, as well as asymptotically valid testing of the original zero-partial-correlation null. Exact finite-sample \(F\)-calibration holds under the projected null. An explicit coordinate-dense Gaussian model verifies the required rates even when the dimension exceeds the sample size. Numerical experiments illustrate the finite-sample performance of the proposed procedures.
Read the full abstract
Combinatorics
Uniquely $K_{4}$-saturated graphs with large maximum degree
A graph $G$ is \textit{uniquely $H$-saturated} if it contains no copy of $H$ as a subgraph, but adding any edge to $G$ creates exactly one copy of $H$ in the resultant graph. In this paper, we consider the uniquely $K_{4}$-saturated graphs and characterize the uniquely $K_{4}$-saturated graphs on $n$ vertices with maximum degree $\Delta\geq n-5$. We prove that if $G$ is a uniquely $K_{4}$-saturated graph with universal vertex $x$, then $G-x$ is either a star or a Moore graph of diameter 2. The uniquely $K_{4}$-saturated graphs with maximum degrees $n-3$ and $n-5$ are $\overline{C_{7}}$ and $G_{10}$ in Figure 2, respectively. But there are no uniquely $K_{4}$-saturated graphs with maximum degree $n-2$ or $n-4$.
Read the full abstract
Signal Processing
Toward Pseudorange-Based PNT with Starlink: A Mock Fused LEO GNSS Implementation
We present experimental results from Starlink for pseudorange-based positioning, navigation, and timing (PNT), emulating a system that fuses a primary communications mission with a secondary low-Earth orbit (LEO) PNT mission. Legacy global navigation satellite system (GNSS) constellations faceinherent vulnerabilities and signal limitations, prompting growing interest in broadband LEO constellations—such as SpaceX's Starlink—for resilient PNT. However, opportunistic use of stand-alone Starlink signals fails to achieve GNSS-competitive accuracy due to frame timing irregularities and reliance on uncorrected poor-quality publicly available ephemeris data. To overcome these limitations, we present a mock implementation of a fused LEO GNSS architecture. Our method extracts time-of-arrival observables from live Starlink Ku-band downlinks and pairs them with high-fidelity, beam-specific orbit and clock models distributed as if provided natively by the constellation operator. In a field campaign, and despite constraints imposed by consumer-grade receiver hardware, the method yields 95% error bounds of 2.33 m horizontally (semi-major axis), 3.37 m vertically, and 11.19 ns in timing, with the actual solution errors falling well within these bounds.
Read the full abstract
Computer Vision
EgoHMP: Achieving Precise Human Motion Prediction in 3D Scenes via Egocentric Cues
The complex interactive dependencies between human behavior and 3D scenes make accurately predicting human intent a major challenge for human motion prediction (HMP). While recent studies use human gaze to infer intent, gaze signals suffer from ambiguity, drift, and a reliance on eye-tracking hardware. Research in cognitive science indicates that egocentric views inherently encode potential interactions, while head motion serves as a crucial precursor in motion planning. Motivated by this, we propose EgoHMP, a novel framework that uses egocentric views and head motions as robust carriers of interaction intent. By translating this intent into an envisioning of future interactions, EgoHMP achieves precise HMP in 3D scenes. EgoHMP first employs a vision-motion alignment encoder to align the visual and motion feature spaces. Based on these aligned features, an egocentric intent predictor with a modality-aware modulation mechanism adaptively suppresses invalid interaction cues to predict an accurate interaction probability map. Furthermore, to enhance the stability and realism of the generated motions, we propose a contact-aware motion decoder. Specifically, inspired by the human cognitive process of planning trajectories prior to action execution, we first predict a global trajectory to serve as a self-prompt, effectively mitigating the optimization conflict in diffusion models. Subsequently, we introduce a multi-stage denoiser, which relieves motion artifacts by incorporating human contact modeling through a multi-stage optimization strategy. Finally, a GCN-based motion decoder is employed to synthesize physically plausible and semantically consistent motions. Extensive experiments demonstrate that EgoHMP achieves state-of-the-art performance.
Read the full abstract
General Topology
On Lévy type subgroups of the symmetric group of $\omega$ associated with simple density functions
The Lévy group, introduced by Lévy in (Lévy (1951)), is defined by \[\mathcal{L} = \left\{\sigma\in aut : \lim_n \frac{|\{k\in n : \sigma(k)\notin n\}|}{n} = 0\right\},\] where $aut$ denotes the group of all permutations of $\omega$. This group happens to be closely connected with the notion of asymptotic density as it consists of permutations that map sets with an asymptotic density back to sets with the same density. Our aim is to replicate the construction by replacing $n$ in the denominator by $g(n)$ for a function $g:\omega\to [0,\infty)$ with $g(n)\to \infty$ (weight), which leads us a class of Lévy type groups. Specifically we define simple Lévy group with weight $g$ as \[\mathcal{L}_g = \left\{\sigma\in aut : \lim_n \frac{|\{k\in n : \sigma(k)\notin n\}|}{g(n)} = 0\right\}.\] As a natural consequence we first study its relationship with the notion of simple density functions (which had been introduced in \cite{gd2} as a general class of density functions by similar manner) paving way for the class of simple density functions and ideals $\mathcal{I}_g$. We have $\mathcal{L}_g = aut$ if and only if $n/g(n)\to 0$ (just like $\mathcal{I}_g = 2^{\omega}$ if and only if $n/g(n)\to 0$). The relation between simple density functions and simple Lévy groups is in expected line. What seems more interesting is the comparative study of these groups. We start with the relation between $\mathcal{L}$ and $\mathcal{L}_g$ and constructing suitable examples it is shown that strict inclusions and incomparability may occur. Finally we come up with an interesting observation that there are continuum many pairwise incomparable simple Lévy groups, each of which is incomparable with $\mathcal{L}$.
Read the full abstract
Human-Computer Interaction
How Well Do Benchmarks Evaluate LLMs' Use for Personal Health Informatics?
Benchmarking is a valuable technique for advancing the capabilities of Large Language Models (LLMs), including in personal health informatics. There are growing efforts on benchmark creation for personal health data activities. However, it is unclear whether and how well these efforts align with needs and considerations of personal health informatics process. Addressing this need, we first analyzed 53,587 tasks across eight benchmarks related to personal health informatics and characterized the space of benchmark design. We then contrasted our observations of design choices with personal informatics theories and empirical knowledge. We find significant mismatches between existing benchmark design and domain guidelines and understanding. This lack of alignment limits the utility of the evidence current benchmarks offer about LLM capabilities. We use domain knowledge to highlight capabilities benchmarks should include and discuss opportunities our work enables for taking a human-centered and domain-driven approach to benchmark design.
Read the full abstract
Artificial Intelligence
DrugCoding: A Scientific Programming Agent for Drug Discovery
Drug discovery involves multiple interconnected research stages, including molecular design, property prediction, and virtual screening. These processes are characterized by complex workflows, interdependent computational tasks, and diverse specialized tools, placing substantial demands on the coordination of scientific models and computational resources. However, existing large language model (LLM)-based scientific agents still face challenges in autonomous planning of complex tasks, adaptive selection of specialized models, and reliable execution of multi-stage scientific computations, limiting their effectiveness in automating complex drug discovery workflows. To address these challenges, we propose DrugCoding, a scientific programming agent system that integrates scientific task planning, adaptive model routing, multi-agent collaboration, and programmatic tool execution. DrugCoding enables autonomous task planning and dynamic model adaptation for complex research objectives. It can flexibly invoke specialized scientific models, autonomously generate and execute code according to task requirements, and leverage multi-agent collaboration to support efficient execution and reliable validation of complex scientific tasks. To evaluate DrugCoding's performance in scientific tool invocation and multi-stage task execution, we conduct multi-turn tool-calling experiments across diverse language models and scientific tool interfaces. Experimental results demonstrate that DrugCoding achieves an average session-level Pass@1 of 97.35% on the full benchmark, outperforming the strongest baseline, Codex, by 9.24%. On the challenging subset, DrugCoding achieves 97.92%, exceeding the strongest baseline.
Read the full abstract
Analysis of PDEs
Classical solvability of the Cauchy problem for multi-order time-fractional systems with full matrices of differential operators
This article investigates the Cauchy problem in the whole space for a system of linear time-fractional partial differential equations with an arbitrary fixed number of components and a full matrix of constant-coefficient spatial differential operators. Each component is governed by its own Caputo derivative, and no rational relation between the fractional orders is required. We assume that the diagonal spatial symbols are homogeneous, positive and elliptic, while the off-diagonal entries have strictly lower order than the diagonal operator in the corresponding column. Using spatial Fourier and temporal Laplace transforms, we construct solution operators through convergent series of time convolutions involving scalar Mittag–Leffler kernels. These series account for all coupling paths, including repeated interactions between components. A fractional Duhamel representation incorporates nonhomogeneous source terms. We establish uniform convergence on bounded frequency regions and weighted multiplier estimates by separating low and high spatial frequencies. Under suitable Sobolev regularity assumptions on the initial data and continuously differentiable source terms, we prove existence and uniqueness of a classical solution, together with continuity and regularity properties in the appropriate Sobolev spaces. The general result includes two-component systems and is illustrated by fully coupled examples.