Graphics
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Thursday 3 September
We present LayoutShop, a novel computational framework for designing creative layouts that frame a given article. Inspired by the actual article layout design process, we enable users to create or select layout templates for conceptualization. These templates help construct a layout design space to extract eligible layout structures. Our algorithm then determines the geometry of the extracted layout structures to frame the given article via an optimization approach. We then employ two neural networks for layout assessment, and the high-quality outputs are returned to users for selection. We conducted a user study to evaluate the framework's usability and the quality of the article layouts it produces. The results of the user study confirmed that our framework can effectively help users create high-quality article layouts.
General Relativity and Quantum Cosmology
Does spatial curvature generate new thermodynamic criticality at the FLRW apparent horizon?
We generalize the apparent-horizon thermodynamic construction of [1] to a FLRW universe with non-zero spatial curvature, non-interacting cold dark matter and holographic-type dark energy. For nonzero curvature, the scale factor is an additional geometric variable in the horizon equation of state. The usual criticality conditions are thus not well defined until a closure prescription for the curvature sector is provided. We introduce a dimensionless curvature variable and restrict the thermodynamic variations to slices of constant value of this variable. For each such slice, a positive holographic coupling and nonlinear powers larger than one guaranty the existence of a unique positive critical point. The critical specific volume, temperature and pressure are shifted by the spatial curvature, while the critical ratio and the mean-field critical exponents remain unchanged. It thus rescales the critical quantities, without generating a new local universality class. We construct a Helmholtz potential out of the physical thermodynamic volume and the entropy conjugate to the rescaled horizon temperature. In the quadratic holographic model, the parametric coexistence curve which is equivalent to the Maxwell construction is derived from the equality of the Gibbs free energies of the competing branches. The associated latent heat disappears at the critical endpoint, resulting in a global first-order coexistence in the fixed-curvature ensemble. Whether a physical FLRW trajectory intersects the critical or coexistence locus is a separate dynamical question because the dimensionless curvature variable generally evolves during cosmological expansion.
Read the full abstract
Optics
Nonlinear Photoemission for Bright Beams in X-Band Photoinjectors
The ongoing development of low-cost and compact x-ray light sources is vital for increasing the number of experiments that can be performed at the atomic scale. One option for a compact and simple photoinjector is to operate in blow-out mode, in which a high-charge, short-bunch length electron beam self-expands by space-charge forces into a uniformly ellipsoidal shape. This mode necessitates a photoinjector whose initial electron pulse has a pulse duration on the order of $100$ fs. A photoinjector utilizing linear photoemission with an ultraviolet laser in this regime presents significant technical challenges, as many optical substrates degrade at these wavelengths. In this paper, we present an alternative design that uses nonlinear photoemission to produce electron beams with high charge and short bunch length in an X-band (9300 MHz) photoinjector. This photoinjector design produces an electron beam with a bunch charge of up to $200$ pC, an RMS bunch length of $291$ fs, a normalized transverse emittance of $1$ to $1.4$ mm$\cdot$mrad, and an energy spread of $10^{-4}$. This results in a beam brightness of $10^{18}$ A/m$^2$, which is bright enough for our beamline's purposes. We also present ongoing efforts are aimed at further increasing the electron beam brightness by using a spatial light modulator to shape the laser profile at the cathode.
Read the full abstract
Software Engineering
Dynamic Adaptation of the LLM Context for Generating Routines with Coupled Semantics
LLM-based code generation fails when correctness depends on execution-dependent coupling: the meaning of one routine is defined by the runtime behavior of another, a relationship that cannot be resolved from textual descriptions alone. This limitation, which we call static binding, is not confined to explicitly coupled problems; it appears to varying degrees whenever correctness depends on joint execution behavior across components, from explicit cross-coupled optimizers to subtler joint constraints in packing, routing, and symbolic search. This paper proposes dynamic context adaptation, a sample-efficient validation-generation loop designed for this setting. A validation agent extracts structured diagnostic information from execution traces, providing gradient-like guidance to a generation agent that proposes multiple candidates per iteration. A knowledge graph derived from the problem description supplies semantic constraints to the generation agent. Simulated annealing selects among candidates to avoid greedy collapse. Our method outperforms zero-shot, Reflexion, and OpenEvolve on seven of eight problems at both 300 and 600 evaluations (p < 0.01), a regime where population-based search has not yet accumulated sufficient diversity to compete. Notably, on the primary motivating problem (cross-coupled optimization), our method also achieves the best score at 1000 evaluations, consistent with the hypothesis that structured execution feedback is most beneficial when correctness depends on runtime coupling. Ablation results confirm that structured execution feedback is the primary driver.
Read the full abstract
Computational Finance
Quantum Circuit Learning for Volatility Modeling: Multifractal Analysis of Realized Volatility Time Series
Herein, we propose a quantum circuit learning framework for modeling the realized volatility (RV) of Bitcoin and investigate the statistical properties of the predicted time series through multifractal analysis. Unlike conventional GARCH-type models, which require a pre-specified functional form for the volatility process, a parameterized quantum circuit directly approximates the volatility function from empirical data, eliminating the need for explicit model selection. Using five-minute Bitcoin price data, we construct daily RV, train a single-qubit parameterized quantum circuit, and generate a long synthetic time series from the optimized quantum circuit. Multifractal Detrended Fluctuation Analysis is applied to calculate the generalized Hurst exponent $h(q)$, the singularity spectrum $f(\alpha)$, and the multifractal scaling exponent $\tau(q)$. The predicted return series exhibits $h(2)\approx 0.5$, consistent with near-random dynamics, and both the predicted and the empirical return series display multifractality that partially persists after random shuffling. The increment series of RV shows pronounced anti-persistence with $h(2)\approx 0.05$–$0.1$, consistent with the rough volatility hypothesis. These results demonstrate that a simple single-qubit parameterized quantum circuit captures qualitatively some observed properties in Bitcoin volatility dynamics.
Read the full abstract
Soft Condensed Matter
A discrete crack-tip theory for nonlinear lattice networks
Crack-tip fields govern deformation localization and failure initiation. Classical continuum fracture mechanics describes these fields through theories such as the Hutchinson-Rice-Rosengren (HRR) field for nonlinear power-law solids. However, continuum descriptions break down near cracks in soft and architected materials, where load is transmitted through discrete chains, fibers, or struts. Here, we develop a discrete crack-tip theory for lattice networks with nonlinear chains. The theory has two central components. First, at large deformation, the strain of a representative chain in layer $i$ depends approximately linearly on the applied macroscopic strain, $\varepsilon_i\approx k_i(\lambda-1)$, defining a layer-dependent strain-amplification factor $k_i$. Second, along topology-selected chain directions, termed discrete HRR lines, the layer-to-layer ratios of $k_i$ follow a two-regime scaling law. Together, for a power-law chain force-strain relation with exponent $p$, the inner discrete regime predicts $\varepsilon_i\sim i^{-1/p}$ and $f_i\sim i^{-1}$, which differs from the classical continuum HRR prediction. The theory also explains why the intrinsic fracture energy approaches a size-independent limit as the network size increases. Photoelastic hydrogel experiments further validate our theory. These results reveal a two-regime crack-tip scaling law in nonlinear lattice networks and provide a framework for predicting chain deformation and intrinsic fracture energy.
Read the full abstract
Quantum Algebra
Even and odd minimal categorifications of the nilpotent part of classical and quantum sl(2)
We write down generators and relations for a minimalist version of the nilHecke category which allows us to categorify the divided powers of a generating functor. We then classify these minimalist counterparts of the nilHecke category, subject to suitable assumptions. A minimalist version of the odd nilHecke category is introduced as well.
Read the full abstract
Cryptography and Security
Optimizing Credential Blast Radius Through Trust Boundaries and Delegation Under Post-Quantum Authentication Costs
Partitioning interacting services into independently rooted trust domains limits issuer-compromise reach while increasing calls across trust boundaries. Post-quantum replacements for public-key authentication and key-establishment mechanisms can increase crossing latency on constrained or lossy paths. We formulate the joint selection of trust domains and credential-derivation structures under policy and latency constraints, linking separate service-interaction and credential-derivation graphs through domain assignment. Credential blast radius measures weighted service impact after compromise. A linear upper bound supports optimization, while a joint event model gives exact expected impact. We identify when risk from issuers trusted across domains can be incorporated into this linear score, avoiding separate issuer-propagation calculations for each candidate. Although the general problem is NP-hard, we identify restricted cases that can be solved efficiently and exactly. Joint optimization yields lower blast radius than choosing boundaries first in 195 of 230 exhaustive synthetic comparisons, especially under chained delegation. A trace-derived replay used measured post-quantum costs, synthetic risk inputs, a fixed derivation family, and one to six trust domains. Under independent compromise events, mean expected impact was up to 36% lower than with one domain within the latency budget. The framework turns risk assumptions and measured crossing costs into candidate trust-domain and credential-derivation designs.
Read the full abstract
Artificial Intelligence
Extremely Sparse Supervision Incentivizes Reasoning Ability
Large language models demonstrate increasingly strong reasoning capabilities through effective post-training. Yet, prevailing post-training methods optimize over massive numbers of tokens, implicitly assuming that effective learning must be token-intensive. We revisit this assumption in the on-policy distillation (OPD) setting, which naturally admits dense teacher supervision at every generated token. Using the Qwen3 family, we discover a counter-intuitive phenomenon: reasoning can be effectively incentivized by an extremely small fraction of generated tokens–as few as one or two tokens per reasoning trajectory, corresponding to only 0.05% of all tokens. Surprisingly, this sparse supervision in most cases matches or surpasses full-token training in improving reasoning ability, despite excluding the vast majority of generated tokens from the training objective. This phenomenon is consistently observed across nine teacher–student configurations spanning different model scales on mathematical reasoning tasks, and is further validated on coding reasoning, Llama models and Proximal Policy Optimization (PPO)-based reinforcement learning with verifiable reward (RLVR). Interestingly, such extremely sparse supervision may be closer to the natural learning process: rather than correcting every step word by word, one reflects on a few critical reasoning steps, updates prior understanding, and continues the trial-and-error, avoiding micro-level corrections while remaining remarkably effective. Overall, our results challenge the assumption that effective post-training must be token-intensive and point to a new direction for understanding and designing more efficient post-training algorithms.
Read the full abstract
Artificial Intelligence
La Agente Óptima: Towards Agentic Self-Driving Laboratories
Self-driving laboratories (SDLs) combine automated experimentation with adaptive decision-making to accelerate scientific discovery. Their operation nevertheless often depends on human specialists who translate scientific objectives into executable closed-loop campaigns. Specialists adjust them as data and operating conditions change. Here, we present La Agente Óptima, an agentic framework that constructs and supervises Bayesian optimization campaigns across computational and experimental systems while maintaining a persistent optimization state. By separating large language model (LLM) reasoning from executed campaigns, Óptima runs repetitive optimization loops consistently, returns control to the agent only when progress requires interpretation or campaign revision, and keeps every decision auditable. We evaluate Óptima across ablation studies, five digital discovery tasks, and two physical platforms. Throughout, Óptima maintained executable campaigns as both the scientific problem and execution environment evolved. In a closed-loop contact angle optimization campaign, Óptima identified and corrected a mid-run measurement failure, bringing the contact angle from 71.4 to 67.8 degrees, just above the 64-66 degree range. From this result, Óptima correctly inferred that the target was likely unattainable with the available reagents and recommended changing the formulation. In a five-day multi-objective flow-chemistry campaign, Óptima increased the yield from 30% to 59% over 23 experiments. Despite substantial inference costs, it cost less and used substantially less starting material than a human-directed campaign, while selecting a more mass-efficient operating point. These results show that LLM-based agents can make rigorous, long-running optimization campaigns accessible to domain scientists without specialist setup, expanding the scope of SDLs.
Read the full abstract
Atmospheric and Oceanic Physics
When a high-mountain slope failure cascades downstream: physical footprint and evolving exposure of the 2026 Gyirong mixed rock-ice cascade
The 26 August 2026 Gyirong cascade connected a high-mountain failure to developed downstream valleys. Event-bracketing observations delineate a preferred changed/source surface of 1.01 km^2, within alternative interpreted envelopes of 0.49-1.84 km^2, and a representative 21.8 km source-to-port route descending 3.40 km. Across two precipitation products, five antecedent windows and six fixed spatial supports, all 60 matched-year ranks remain below the 90th-percentile wet threshold. The source-nearest 7-day temperature mean of 9.43 degrees C exceeds all 25 matched years from 2001-2025. Positive temperature anomalies extend to every tested support, but ranks vary from the 88th percentile to above all historical values. Within the subsequently affected 37.354 km^2 UNOSAT footprint, modelled built-up surface increases from 0.036 to 0.441 km^2 between 1975 and 2020; its fraction of the fixed area rises from 0.096% to 1.181%. Delivered rapid maps contain 695 building points and 15.671 km of roads graded destroyed. These measurements connect the physical footprint with environmental context and historical exposure development. They complement existing process reconstruction while leaving the initiation mechanism and the mechanical role of warming unresolved.
Read the full abstract
Earth and Planetary Astrophysics
A Cross-Laboratory Comparison Study of Titan Haze Analogs: Surface Energy II
The organic haze particles on Titan play important roles in atmospheric cloud formation and aerosol-lake interactions. These processes are strongly influenced by the surface energy of the haze particles, which controls cohesion and wetting behavior. This study presents a comparative analysis of 32 laboratory-produced haze analog samples ("tholins") synthesized across three laboratories. Using contact angle measurements, we determine the total surface energy and its dispersive and polar components for all samples, systematically evaluating the effects of substrate choice, air exposure, initial N2/CH4 gas mixture, and experimental setup. We find that tholin samples exhibit minimal substrate dependence, whereas exposure to ambient air substantially modifies the surface chemistry, altering the balance between dispersive and polar components. Thus, future Titan-relevant surface property measurements may use any substrate but must keep samples pristine. Surface energies vary weakly across methane concentrations, from 1-10% CH4 in N2, indicating that Titan's hazes formed at different altitudes should exhibit broadly similar cohesiveness. In contrast, experimental conditions such as gas exposure time and energy source produce the dominant differences in surface energy, driven largely by variations in polar components. Despite these differences, tholin samples exhibit high dispersive components, implying that Titan's hazes should act as efficient cloud condensation nuclei for hydrocarbon clouds and should generally sink into Titan's lakes. Given observed ethane ice clouds, we conclude that cold plasma tholins may be better physical analogs for Titan's hazes than samples produced with far-ultraviolet irradiation, though intrinsic surface energies of UV tholins remain uncertain due to film thickness limitations.
Read the full abstract
Artificial Intelligence
Reducing Hallucinated Transcripts in Whisper via Hallucination Space Projection
Whisper is a widely used foundation model for automatic speech recognition (ASR), but its generative decoder can produce fluent hallucinated transcripts for inputs containing little or no speech. We propose a training-free, inference-time method to reduce these hallucinations using low-rank projection of decoder activations. A compact hallucination-associated subspace is estimated from non-speech calibration data, and decoder hidden states are projected away from this subspace during inference. We evaluate two variants: always-on, which applies projection to all inputs, and gated, which applies it only when Whisper predicts that an input is likely non-speech. Across non-speech benchmarks, always-on projection reduces average hallucination rate (HR) from 31.31% to 2.44%, a 92.21% relative reduction, while gated projection reduces HR to 3.74%, an 88.05% relative reduction, with lower false rejection of genuine speech. On LibriSpeech, gated projection increases absolute word error rate (WER) by 0.33-4.39 percentage points and yields false-rejection rates (FRR) of 0.41–9.97% across model and split settings. These results show that low-rank activation projection can substantially suppress Whisper hallucinations without retraining, while providing a controllable trade-off between hallucination suppression and speech recognition performance.
Read the full abstract
Astrophysics of Galaxies
Star Formation Evolution in Galaxy Pairs: Constraints from Morphological Disturbances and Recent Star Formation Histories
Galaxy interactions can enhance star formation, but how star formation evolves during galaxy interactions remains poorly constrained by observations. We combine pair separation, morphological disturbance, and recent star formation history to study this evolution. We measure morphological disturbance with the shape asymmetry parameter ($A_{\rm shape}$) using deep images from the DESI Legacy Surveys, and use Galaxy Zoo DESI classifications as an independent test. We infer recent star formation histories by comparing SFR and $\mathrm{EW}(\mathrm{H}\alpha)$, which trace current star formation, with $\mathrm{EW}(\mathrm{H}\delta_A)$ and $D_n4000$, which are sensitive to stellar populations formed over longer timescales. At small projected separations, strongly disturbed galaxies show the strongest enhancements in SFR and $\mathrm{EW}(\mathrm{H}\alpha)$, indicating that close encounters trigger strong star formation. At intermediate separations ($d_{\rm p}\sim100\,\mathrm{kpc}$), their current star formation is only moderately enhanced, but their $\mathrm{EW}(\mathrm{H}\delta_A)$ enhancement is the strongest. This indicates a larger contribution from intermediate-age stars formed during stronger star formation in the past $\sim0.1$–$1\,\mathrm{Gyr}$. The TNG100 analysis shows that this pattern results from rapid changes in SFR around close passage. SFR rises sharply during the encounter and declines as the galaxies move apart, while tidal disturbances remain visible. As short-lived massive stars disappear and intermediate-age A-type stars begin to dominate the spectrum, the $\mathrm{EW}(\mathrm{H}\delta_A)$ enhancement peaks later than the SFR enhancement. Our results provide important observational evidence for the evolution of SFR along the merger sequence.
Read the full abstract
Artificial Intelligence
IPGeoAI: Transformer-Based Geolocation with LLM Semantic Fusion
Accurate city-level IP Geolocation is an important enabler for the modern digital ecosystem, underpinning services ranging from local content delivery and targeting to digital rights enforcement. However, traditional heuristic and database-driven methods often struggle to resolve the complex, non-linear allocation patterns of modern network infrastructures, particularly within the exploding IPv6 address space and transient mobile networks. In this paper, we introduce IPGeoAI, a novel deep learning model architecture that reframes geolocation from a static lookup problem to a sequential modeling task. Our approach utilizes the Transformer Encoder to capture hierarchical dependencies inherent in IP subnet structures. We propose a method to resolve geographic ambiguity by integrating unstructured semantic context via a Zero-Shot LLM Feature Extraction pipeline. We utilize Large Language Models to transform raw, noisy Autonomous Systems (AS) descriptions into structured, domain-specific metadata (such as 'University' vs. 'ISP' or 'Global' vs. 'Local') via an offline pre-computation process. By fusing these semantic signals into the network via a Multi-Head Cross-Attention module, we bridge the gap between numerical network topology and real-world semantic identity. Extensive offline evaluation on a proprietary dataset spanning 200,000 cities demonstrates that IPGeoAI significantly outperforms a leading external vendor in city-level granularity. By adopting a hierarchical inference strategy that refines coarse-grained country signals, our model achieves a 6% improvement in city-level accuracy while extending coverage to 100% of the traffic. Furthermore, in large-scale online production tests, the model drove a statistically significant +0.35% improvement in our 1st-tier downstream use cases metric.
Read the full abstract
Materials Science
Structural tuning of reduced exciton mass in layered HOIP compounds: Causation vs. correlation
Reduced exciton mass ($\mu$) was recently reported to correlate strongly with a framework distortion in a series of nine single-layer (2D) metal-halide perovskite (HOIP) compounds. Specifically, $\mu$ was observed to increase in tandem with an alternating PbI4 octahedral tilt about an in-plane axis. In this work, we use group representation theory to decompose the observed framework distortions into displacive symmetry modes of a common high-symmetry parent framework. We find that all nine distorted frameworks involve linear combinations of the same six symmetry modes, which have been reported to contribute to the framework distortions of a wide range of HOIP compounds. We show that these modes have highly correlated impacts on the band structure. To differentiate causation from correlation, we vary the amplitude of each mode independently and use density-functional theory to determine the resulting electronic band structures, from which $\mu$ is extracted. We find that bond-transverse displacements of the equatorial halide atoms increase $\mu$, while bond-transverse displacements of the apical halide atoms decrease it. Bond-axis displacements appear to have little or no effect on $\mu$. Our results demonstrate three new structure-property relationships, revealing a promising new avenue for exciton engineering in layered perovskite materials.
Read the full abstract
Quantum Physics
A Sim-to-Real Study of Surface-Code Decoder Benchmarking
Quantum error-correction decoders are typically benchmarked against synthetic circuit-level noise, under the assumption that a decoder's ranking under such noise transfers to hardware and improves as the noise model becomes more realistic. The Willow processor, the first to operate below the surface-code threshold, allows us to test this assumption. We rank a panel of six decoders using a four-rung ladder of noise models with increasing fidelity, evaluated against real data across three code distances, two bases, and fifteen round counts. Rank agreement with hardware appears once the noise model gives each operation type its own error rate. Calibrating the model to the device improves absolute error rates but not rank agreement. We additionally provide the first independent evaluation of NVIDIA's Ising pre-decoder on hardware, at code distances below its training receptive field and via a mapping onto the lattice on which it was trained. Under these conditions, it holds no accuracy-latency advantage: another panel decoder matches or improves on it in both per-cycle error rate and decode latency in 278 of the 280 evaluations. We release the full pipeline and the per-shot outcome of every evaluation, so future decoders and devices can be compared.
Read the full abstract
Computation and Language
Rhythms of Work: Multi-Scale Interpretation of Human Behavioral Traces for Workplace Agents
Runtime traces are becoming a central substrate for understanding agentic systems, yet interpretation has focused largely on what the agent did. Workplace agents face the complementary problem: interpreting the human activity that surrounds them. Hours of low-level events carry rich evidence about a user's state but are too granular to reason over directly, and flattening them into one stream or compressing them into a single embedding both treat "summarize the user's behavior" as if it had one correct answer. We argue instead that behavioral interpretation is resolution-dependent: the same trace should admit multiple addressable interpretations at different temporal resolutions. We construct a multi-resolution vocabulary of semantically normalized operators, recurring motifs, coherent episodes, and day-level rhythms, each preserving the structure salient at its own horizon. Applied to 667 million human-attributed events from a large commercial productivity suite (50,000 users, 100 organizations), it yields 120 operator types, thousands of motifs, 25 episode types, and five day-rhythm archetypes. We validate it on real telemetry: re-running the entire pipeline on a disjoint 2,000-user sample recovers the same taxonomy (structural stability), and on held-out users the full representation forecasts a user's next episode more accurately than a flat-operator baseline, a 17% relative macro-F1 gain (predictive validity), so the abstractions preserve future-relevant information rather than merely describe it. A controlled resolution ablation then shows that no single level is optimal across questions: different agent-facing questions about the same trace are best answered at different resolutions. Behavioral trace interpretation for agents should therefore be multi-resolution and query-conditioned: an agent should access the temporal grain a question needs, not one universal summary.
Read the full abstract
Computer Vision
DART: Depth-as-Target Pretraining for Surgical Vision Foundation Models
Vision foundation models (VFMs) are valuable in data-scarce domains such as surgery, where a single pretrained backbone can provide rich representations for many downstream tasks. Yet the dominant self-supervised pretraining paradigm uses only RGB images, leaving readily available complementary signals, such as depth maps, unused. This is a particular missed opportunity in surgery, where natural-image VFMs transfer poorly while the scene geometry is rich and informative. With strong off-the-shelf models now able to produce pseudo-labeled dense depth for any image corpus, we hypothesize that such signals can be folded into pretraining to learn better representations. We present DART, an RGB-D pretraining recipe that builds on DINOv2 with a simple modification: a pixel-space depth reconstruction objective applied to masked iBOT patches, supervised by pseudo-labeled depth. Depth is used only during pretraining, so fine-tuning and inference remain RGB-only. We find that this pixel-level reconstruction head improves representation quality rather than disrupting it. We further show that depth, which encodes scene geometry, is more effective as a target than alternative dense signals such as Canny edges, confirming that the gains stem from depth rather than added supervision alone. Across eight surgical benchmarks spanning segmentation, depth estimation, and image-level recognition, DART outperforms both natural-image and in-domain baselines, including a vanilla DINOv2 trained on identical data, improving dense prediction while also strengthening image-level understanding. More broadly, DART shows that freely available geometric pseudo-labels can strengthen foundation model pretraining without extra labels or added inference cost, pointing toward stronger backbones for surgery.
Read the full abstract
Combinatorics
Universality in the algebra and topology of cographs
A finite simple graph $G$ is called a cograph if it does not contain the path on four vertices $P_4$ as an induced subgraph. It is classically known that the family of cographs are well-quasi-ordered by the induced subgraph relation \cite{D}. In preceding work of Knudsen and the third author \cite[Theorem 7.2]{KR}, it was shown that this well-quasi-order statement admitted a categorification, which allowed those authors to prove universal finite generation statements about the homology groups of configuration spaces on cographs \cite[Theorem 1.5]{KR}. In this work, we expand \cite[Theorem 7.2]{KR} to be compatible with the family of polynomial rings on the vertex sets of cographs. By consequence, we are able to prove a number of universality results related with edge and toric ideals of these polynomial ring, partially generalizing and expanding upon work of Kahle \cite{kahle2019binomial}. We also conclude strong restrictions on the kinds of topologies that can arise from graph complexes and anchored configuration spaces associated to cographs, as well as combinatorial constraints on the possible combinatorics of hyperplane arrangements of cographs.
Read the full abstract
Atomic and Molecular Clusters
X-ray photoelectron spectroscopy of Ar and Kr clusters formed in He nanodroplets
We report the first soft x-ray photoelectron spectroscopy (XPS) measurements of Ar and Kr clusters formed inside superfluid helium nanodroplets (HNDs) through consecutive pickup of dopant atoms. Ar and Kr atoms and clusters are selectively inner-shell ionized (Ar 2p, Kr 3d) using monochromatic soft x-ray synchrotron radiation. In the regime of strong doping, the electron spectra exhibit features characteristic of free Ar and Kr atoms as well as their clusters. The Kr cluster spectra agree well with literature data for bare Kr clusters. Backed by detailed simulations of the pickup process, this agreement indicates that the observed spectra originate from nearly bare Ar and Kr clusters, from which most or all He has evaporated in the course of cluster aggregation. These results establish HNDs as a platform for XPS of various types of molecular complexes and nanostructures.
Read the full abstract
Robotics
Continual Field-Adaptive Models (CFAMs) for Post-Deployment Physical AI
Unattended interactive autonomy - machines that step into danger in place of humans and complete tasks with human tools - remains a missing capability in mission-critical operations. These domains offer scarce training data and only onboard compute, yet deployed systems must face novelty without erasing prior competence. We introduce Continual Field-Adaptive Models (CFAMs), which learn efficiently in the lab and continue learning after deployment through autonomous, gradient-free, on-device updates. CFAM uses a complementary learning architecture with a frozen slow-learning component and a fast-learning Capsule Field. The slow component contains three cortices: Sensor, which maps multimodal input into 3D-grounded geometry; Reasoning, which decomposes tasks into skills and evaluates outcomes; and Action, which executes geometric skills. The Capsule Field stores field learning one-shot and gradient-free as Competence Capsules. Skill installation is few-shot in the lab and continual in the field; open-world novelty is outside scope. We evaluate CFAM across five embodiments: manipulator, quadruped, humanoid, quadrotor, and off-road vehicle. Baselines (pi0, CogACT, SpatialVLA) use the same in-house multi-embodiment dataset for physical-platform comparisons. CFAM reaches the operating point of a standard policy trained on the full prior-training dataset using 40% of the data, or 2.5x fewer trajectories. At test time, autonomous capture of verified near-OOD cases improves action success by 13.9 percentage points. In sequential simulation, backward transfer is -0.5 percentage points versus -11.4 for LoRA. CFAM therefore provides a bounded form of post-deployment physical intelligence: few-shot skill learning, autonomous field growth from verified near-OOD experience, and retention of prior competence.
Read the full abstract
Number Theory
Hilbert's Irreducibility for $\mathbb{G}_m$
Let $K$ be a number field and $S$ a finite set of non-archimedean places. Write $\mathcal{O}_S$ for the ring of $S$-integers of $K$ and $\mathcal{O}_S^\times$ for its unit group. Let $\pi : X \rightarrow \mathbb{P}^1$ be a morphism of (irreducible) curves defined over $K$, and denote by $\operatorname{Red}(\pi)$ the set of $\alpha \in \mathbb{P}^1(K)$ such that the fibre $\pi^{-1}(\alpha)$ is reducible (i.e. the Galois action on the fibre is intransitive). Hilbert's Irreducibility Theorem asserts that $\operatorname{Red}(\pi)$ is contained in a thin subset of $\mathbb{P}^1(K)$. In this paper we give an explicit description of $\mathcal{O}_S^\times \cap \operatorname{Red}(\pi)$. As an application we prove the following result inspired by a classical theorem of Pólya and Siegel. Let $p_1,\dotsc,p_s$ be rational primes. Let $f \in \mathbb{Q}[x]$.Then the following are equivalent: - There are infinitely many tuples $(e_1,\dotsc,e_s) \in \mathbb{N}^s$ such that the polynomial $f(x)-p_1^{e_1} \cdots p_s^{e_s}$ is reducible. - $f=p_1^{a_1} \cdots p_s^{a_s} g^\ell$ (with $\ell$ prime) or $f=-4 p_1^{a_1} \cdots p_s^{a_s} g^4$ for some $g \in \mathbb{Q}[x]$ and some integers $a_1,\dotsc,a_s$.
Read the full abstract
Computer Vision
VISTA: Dense Multi-Label Classroom Coding with Vision-Language Models
Video-language benchmarks are usually constructed by the dataset authors without published reliability statistics, leaving the noise floor of the construct unknown. We argue that multimodal benchmarking benefits from methods taken from research communities that have already invested in strategies to ensure reliability. We illustrate the case with the Classroom Observation Protocol for Undergraduate STEM (COPUS): a 24-code multi-label observation instrument with a decade of peer-reviewed reliability literature. We recast COPUS as a video benchmark for multimodal foundation models, where it provides a dense set of structured labels (a 24-dimensional binary vector every 2 minutes across a 50-90 minute lecture), an externally validated vocabulary, and established literature that provides a per-code reliability target based on human evaluators. Annotations in our evaluation corpus are produced by a 5-person human-evaluator panel whose consensus matrix is our reference. We propose VISTA, a baseline that runs MiniCPM-V-4.5 over a dense sliding window, refines its per-window outputs with a lightweight multi-layer perceptron (MLP) head trained on top of the frozen backbone, and max-pools the resulting predictions onto the 2-minute COPUS grid. On three held-out chemistry lectures, VISTA reaches 80.1% restricted macro accuracy versus 74.9% for the zero-shot variant, with the largest residual errors on visually similar instructor codes and on rare audio-dependent codes. We characterize three systematic failure modes (audio-partial observability, fine-grained group-work discrimination, long-tail recall) and release the benchmark tooling, prompts and baseline code at https://github.com/ajfranck/VISTA.
Read the full abstract
Machine Learning
Fast Surrogate Modeling of Excitable and Oscillatory FitzHugh-Nagumo Dynamics with Parametric Neural Operators
The FitzHugh-Nagumo (FHN) system serves as a simplified model of neuronal voltage dynamics, capturing the activator-inhibitor structure behind both isolated action potentials and the rhythmic spiking seen across the brain. Exploring its 5D physiological parameter space is important for neuromodulation and mapping voltage recordings back to biophysics, yet classical finite-difference solvers make rapid parameter sweeps expensive. We train parameter-conditioned Fourier Neural Operators (FNOs) as fast, differentiable surrogates for the FHN voltage and recovery fields on a one-dimensional spatial domain, conditioning each Fourier layer on the parameter vector $\lambda = (D_u, D_v, a, b, \tau)$ via feature-wise linear modulation (FiLM). We apply a single bifurcation analysis that delimits the two distinct regimes the model spans, oscillatory (tonic firing) and excitable (action-potential propagation), and we train one operator in each. In the oscillatory regime the surrogate attains sub-$0.1\%$ relative $L^2$ error on both fields, runs nearly three orders of magnitude faster than the finite-difference baseline, generalizes uniformly across the parameter space, and extrapolates to low single-digit percentage errors outside of the training bounds. In the excitable regime the same operator accurately reproduces the firing threshold and the $c \propto \sqrt{D_u}$ conduction-velocity law and replicates full traveling pulses, fully capturing the excitable bifurcation structure rather than just smoothly interpolating fields.
Read the full abstract
Soft Condensed Matter
From Single-Chain Dynamics to Structure Formation: Dynamic Self-Consistent Field Theory and Molecular Dynamics of (Co)polymer Melts across Entanglement Regimes
Dynamic self-consistent field theory (DSCFT) provides an efficient continuum framework for studying structure formation in inhomogeneous polymer systems, but its predictive accuracy depends on the choice of the nonlocal mobilities. Here, we construct mobility functions for moderately and strongly entangled homopolymer and diblock copolymer systems from the relaxation dynamics of single-chain structure factors, based on molecular dynamics (MD) simulations of the Kremer-Grest model and analytical reptation theory. Single- chain mobilities are combined such that the resulting DSCFT accounts for the dependence of fluxes on local chain densities. The theory is then applied to the spinodal decomposition of symmetric homopolymer blends and diblock copolymer melts following a quench into the (micro)phase-separation regime. Predictions of DSCFT are systematically compared with MD simulations. Mobility functions derived from single-chain dynamics are found to reproduce the kinetics of structure formation more accurately than conventional Debye-type mobilities. We additionally investigate the influence of adding stochastic currents (noise) that are correlated according to the fluctuation-dissipation relation. At low noise levels, they enable the generation of equilibrium initial states and facilitate defect annealing. At high noise levels, however, nonolinear effects lead to discrepancies between DSCFT and MD simulations.
Read the full abstract
Distributed Computing
Revisiting MemGuard Overhead: A Reproduction Report
As an increasing number of embedded platforms incorporate multiple processing units, shared resource contention induced unpredictable execution time poses a challenge for real-time system design. Memory bandwidth regulation is a popular mitigation approach, and MemGuard is the canonical example. Recently, MemPol introduced a new bandwidth regulation mechanism, which was compared with MemGuard. Specifically, they reported significant overheads for MemGuard, citing up to a 1.79x slowdown, to contextualize MemPol's comparative benefits. This report is meant to clarify and add the necessary nuance to the experiments carried out in that prior work. Specifically, we show that the MemGuard overheads presented in these prior evaluations were unintentionally amplified as the result of using a suboptimal configuration with an older version of MemGuard, wherein the benchmark under test was pinned directly to the master core responsible for handling global timer interrupts. By faithfully reproducing these specific experiments using a modern, decentralized implementation of MemGuard, we demonstrate that the actual execution overhead drops significantly under identical conditions. Consequently, when evaluated with a properly configured recent version, MemGuard exhibits an overhead that is highly comparable to MemPol's overhead. By revisiting these baseline metrics, this report provides an updated and comprehensive perspective required for future evaluations of memory bandwidth regulators.
Read the full abstract
Plasma Physics
Assessing the effect of error field penetration during plasma current ramp-up in the DIII-D tokamak
This work provides evidence that established error field penetration threshold scalings remain applicable during plasma current ramp-up. In dedicated DIII-D experiments with imposed $n=1$ perturbations during extended $I_p$ ramps, an apparent empirical threshold is found between $2$ and $3$~kA of applied 3D coil current, above which MHD modes are seeded. The imposed perturbation couples to the rational surfaces present during the ramp, seeding near the $q=4$ surface and penetrating as an $m/n=3/1$ mode by the end of the perturbation phase. To interpret these observations, multi-machine penetration threshold scalings are combined with equilibrium-based overlap metrics from the GPEC code, including the in-situ error fields of the device. This modeling reproduces the observed onset in the amplitude scan and classifies mode seeding across a database of 12 ramp-up discharges spanning a range of plasma currents and densities. Across this database, the seeding appears to be controlled primarily by the applied 3D coil current rather than by the plasma current or its ramp rate. Accounting for the in-situ error fields is found to be important for reliable prediction. These results are consistent with the robustness of scaling-based penetration metrics when coupled to detailed 3D field modeling under transient ramp-up conditions, and suggest the importance of accounting for in-situ error fields when assessing additional externally induced perturbations. This work is motivated by future tokamaks in which transient, non-axisymmetric error fields can arise during startup, for example from runaway electron mitigation coils.
Read the full abstract
Robotics
SocioGesture: Real-Time and Adaptive Social Gesture Perception for Human-Robot Interaction
Robots interacting with people must recognize not only explicit commands, but also social cues such as invitations, refusals, and unavailability. In real deployments, these cues must be inferred from noisy onboard perception under partial occlusion, changing viewpoints, and strict latency constraints. We present SocioGesture, a real-time adaptive social gesture perception system for human-robot interaction (HRI). SocioGesture uses a compact confidence-aware body-hand skeleton representation and a lightweight dual-stream model that fuses body motion with hand articulation for low-latency onboard recognition. To improve deployment robustness, we train the model with occlusion-aware skeleton corruption, exposing it to missing hands, occluded arms, and temporally unstable keypoints without increasing the inference cost. On a social gesture dataset collected in mixed indoor-outdoor HRI scenarios, SocioGesture achieves strong held-out-subject recognition, substantially improves robustness under structured joint occlusion, and runs in real time on a robot-mounted edge device. During deployment, uncertain interaction segments are saved for offline labeling and adaptation, enabling SocioGesture to expand its gesture vocabulary while preserving performance in the original classes. These results demonstrate a practical path toward robust, efficient, and adaptive social perception for interactive robots.
Read the full abstract
Systems and Control
Integrating Agentic Artificial Intelligence with High-Performance Computing for Grid Planning
We present AgentiGrid, an agentic artificial intelligence (AI) framework that integrates large language models (LLMs) intelligence and high-performance computing (HPC) to streamline and accelerate the multi-scenario power flow studies. AgentiGrid is an autonomous decision-making agent that proposes parameter modifications, invokes analyses through HPC analysis toolkit ExaGO, interprets results, and determines subsequent actions. ExaGO provides multiple power flow applications that can perform deterministic, stochastic and security constrained optimal power flow analyses. AgentiGrid provides backends to multiple LLMs (OpenAI, Anthropic, Ollama, and Ollama cloud) augmented with context specific and task specific prompts. Key features include interactive mid-search steering, goal-type-aware post-search analysis, and concurrent variant exploration for power flow optimization. A Streamlit-based graphical launcher provides real-time visualization of iteration progress and generates reports in natural language. AgentiGrid is capable of autonomously converging transmission constrained alternating current optimal power flow in under 20 iterations, with near-perfect reliability