Computer Vision
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Monday 9 November
When a mesh reconstructed from aerial remote-sensing imagery disagrees with reference geometry, deciding what to change requires a diagnosis of the reconstruction process. We present GeoPilot, a workflow that turns diagnostics and prior observations into executable program or camera-configuration interventions. A large language model proposes the change, numerical tools reconstruct the mesh, and an external evaluator measures its geometry under a stated coordinate and reference protocol. The main case tests a shared-intrinsics hypothesis proposed before the comparison. On 224 Metashape-raster development images, shared intrinsics improved all six native-coordinate measures and reduced mean surface-to-reference distance by 7.04%. Reference-assisted alignment fitted and evaluated on separate within-scene checkerboard samples reversed the mean-distance ranking, limiting that gain to native-frame delivery. A separate matched comparison on publisher-undistorted images favored fixed published cameras on all six measures, reducing mean distance by 12.11%; the cameras retain upstream image and laser-scanning adjustment information. With one reconstruction per configuration, these comparisons provide conditional evidence for camera handling. Three end-to-end runs each delivered a mesh and six-metric evaluation from all 327 images of a second development scene.
Rings and Algebras
Becker's Conjecture on Compositions of Higher Degree Diagonal Forms
Becker's conjecture asserts that a rational composition of sums of $d$-th powers of size $[r,s,n]$, with $d>2$, requires $n \ge rs$. We show that this formulation fails over every characteristic-zero field in odd degree, and in every degree over a suitable finite extension of $\Q$. Over formally real fields in even degree, however, we prove that every polynomial composition satisfies $n\ge rs$.
Read the full abstract
High Energy Physics (Phenomenology)
Origin of Tensor Polarized Structure of the Deuteron
The microscopic origin of the deuteron tensor polarized structure function $b_1$ remains unresolved. Conventional nuclear models attribute a nonzero $b_1$ mainly to the $D$-wave component of the deuteron wave function, but generally fail to reproduce the magnitude and momentum-dependence observed by the HERMES experiment. Using a QCD-based light-front framework with degrees of freedom beyond the conventional nucleonic picture, in particular including hidden-color components and relativistic spin–momentum correlations, we find that a small but nonzero estimated hidden-color admixture can be sufficient to generate a substantial tensor response ($b_1(x)$), while it is essentially insensitive to the $D$-wave component. The physical $D$-wave probability of $0.85\%$, fixed by the empirical quadrupole moment, changes $b_1$ by only $1.7\%$, with the dominant contribution arising from the colored-cluster $S$-wave sector. These results demonstrate that the tensor-polarized structure of the deuteron is driven primarily by hidden-color dynamics rather than by the $D$-wave component.
Read the full abstract
Quantum Physics
Model-family selection and recovery-gate design in generator-based noise tomography
Family selection in generator-model noise tomography is a hypothesis test whose shot floor has closed form under a purely coherent noise model, $N^{*}=\lambda_{\mathrm{req}}/(r_{\mathrm{b}} S)^{2}$: set by the absolute bias $r_{\mathrm{b}} S$, not the relative residual, and computable from a calibration alone. The same computation prices the recovery gates, fingerprint and Fisher information coming off the same map, so there is no design-independent verdict and no family-independent design optimum. Whether a design can move that price is structural: a bias confined to the block every design measures identically is design-blind, one that reaches the code blocks is repriced by a factor of sixty-eight—from a dominated design to the cheapest, not between two optima—and the two platforms studied here are its extremes. At the reference point the floor is $1.1\times10^{7}$ and $6.7\times10^{6}$ shots per component, the easy corner of the sampled working points, whose median floor is eighty-three times dearer in one direction and less than twice in the other; at the $10^{5}$ shots a run spends, the power equals the level and the residual reported is shot noise. The design problem is conditioning, not rank: across the sixty-four placements the rms varies by factors of $2.7$ and $3.9$, while the optimal gate count is flat at two to four. At five qubits the whole sample moves to $10^{9}$–$2\times10^{10}$ shots, a design spread of twenty-one against nine at three qubits; on a measured three-qubit register the floor is a requirement on systematic control before it is a shot count, and where that control is missing no budget supplies it.
Read the full abstract
Information Theory
Bounds and Constructions for Optimal Wide-Gap Frequency Hopping Sequences
In a frequency-hopping (FH) system, to reduce interference to the received signal, each user's frequency hopping sequence (FHS) shall have wide-gap property. In this paper, we update the upper bounds on the minimum gaps of optimal uniform wide-gap frequency hopping sequences (WGFHS) with parameters $(v,s,\lambda;e)$ where $v=ks, k\ge 3, \lambda=k$ and $s\ge k(k-1)+1$ is odd. We completely prove the existence of such WGFHSs for $k\le 3$.
Read the full abstract
Group Theory
Group cosets with all elements automorphically equivalent
Let $G$ be a finite group and let $N$ be a proper non-trivial normal subgroup of $G$. We say that $(G,N)$ is an automorphically equivalent pair if for every element $x\in G\setminus N$, the elements in the coset $xN$ belong to the same orbit under the action of ${\rm Aut}(G)$. This generalizes the concept of a Camina pair and constitutes a particular type of equal order pair. In this note, we determine all automorphically equivalent pairs in the class of finite abelian groups.
Read the full abstract
Artificial Intelligence
Reproducible quantum clustering for single-cell transcriptome atlases
Current single-cell clustering pipelines discard most genes and can yield inconsistent clusters across random seeds. We introduce KQE-2, a training-free hybrid quantum-classical framework for reproducible whole-transcriptome clustering. KQE-2 maps cells into entanglement-free quantum states, enabling exact feature evaluation on classical hardware and direct checks on a quantum processor. Theoretically, we prove that removing entanglement is necessary and sufficient to bound single-qubit feature variance locally. Practically, while retraining scVI shifted 14% of cells and altered rare-type markers and drug targets, KQE-2 delivered identical biology across runs. On standard highly variable genes, the quantum encoding alone raised accuracy on a developing cortex from 0.61 to 0.79. KQE-2 scales to 11.4 million nuclei on one server, outperforming the compared methods on eight of ten complete atlases (and within the best method's run-to-run range on two), and leads on six even under standard downstream clustering.
Read the full abstract
Machine Learning (Statistics)
Breaking the $\sqrt{d}$ Communication Barrier in Federated Sampling with Adaptive Hamiltonian Monte Carlo
This work proposes the Adaptive Federated Hamiltonian Monte Carlo (AFHMC) for Bayesian federated learning (FL) tasks. The prior work on HMC applications for FL achieves $O(\sqrt{d}/\epsilon)$ communication cost under log-convex distributions, already demonstrating its advantage over the Langevin dynamic-based counterpart. Based on a refined analysis, this paper further reveals the benefit of HMC for the FL regime by incorporating a control mechanism for local node shifts. We prove that AFHMC at most requires $O(\{d/\epsilon^2\}^{1/3})$ communication cost up to a logarithmic term, which is significantly better than the existing results. Moreover, the proposed method adapts to client heterogeneity. Under low-heterogeneity settings (i.e., the target sampling precision $\epsilon^2$ is higher than the heterogeneity level), the communication rate of AFHMC can be as good as $O(\log(d/\epsilon^2))$, resembling the communication costs for FL optimization tools. To the best of our knowledge, this provides the first logarithmic communication complexity result for federated sampling, even though it requires low heterogeneity. Numerically, we also justify AFHMC's advantage under the low-heterogeneity regime. Our code is available at https://github.com/laaaarrywang/Adaptive-Federated-HMC
Read the full abstract
Statistics Applications
A Review of Measurement System Uncertainty Handoffs Across the Manufacturing Decision Lifecycle
Measurement systems generate uncertain information that downstream manufacturing decisions often consume as if fixed. This structured integrative review examines how measurement error, uncertainty, GR&R estimates, calibration evidence, and drift information are handed off into capability inference, retrospective Phase I reference estimation, prospective Phase II monitoring, operational action, and requalification. Studies are compared by the upstream uncertain object, whether it is known, calibrated, or estimated, its propagation mechanism, the downstream decision, and the validation boundary. The synthesis distinguishes recorded production data from latent-process targets: measurement-system information calibrates limits, uncertainty statements, replication, guard bands, or losses, but does not reveal the realized random error in an individual record. Within the retained literature, measurement-to-capability and measurement-to-monitoring mechanisms are more extensively represented by model-based studies, whereas prospective evidence is thinner where statistical results must support source attribution, cost-sensitive action, and requalification. A decision-handoff spine separates measurement generation, statistical inference, and operational validation so that capability, stability, attribution, and intervention effectiveness are not conflated. The review is integrative rather than exhaustive. Its contribution is an auditable measurement-science framework that preserves provenance, information status, validity conditions, and downstream use while identifying claims that are model-supported, conditionally calibrated, or not yet operationally validated.
Read the full abstract
Classical Analysis and ODEs
The solution of the Bochner–Krall problem
We solve the Bochner–Krall problem posed by Krall in 1938, extending the classification beyond regular orthogonal sequences. We classify monic polynomial sequences for which multiplying each polynomial by the variable gives a linear combination of that polynomial and the polynomials of the immediately preceding and following degrees, and whose members are eigenpolynomials of one fixed nonscalar differential operator of arbitrary finite order with polynomial coefficients. Orthogonality is not assumed. The general formulation allows sequences to start at any fixed degree. In every degree beyond a fixed threshold, each sequence agrees, up to an invertible affine change of variable and monic normalization, with a Hermite, Laguerre, Jacobi or Bessel family, the monomial basis, or a finite rational Darboux transform of a Laguerre or Jacobi family. In the positive-definite case, we determine the representing measures, prove their uniqueness and specify the admissible endpoint masses. No boundedness assumption on the support or asymptotic conditions on the recurrence coefficients are imposed.
Read the full abstract
Computers and Society
Silinx: An Open, Browser-Based HDL Design Platform for Teaching Digital Systems on FPGA Boards
Many digital systems courses teach on FPGA boards whose vendor tool is no longer maintained: the Digilent Basys2 and Nexys2 (Xilinx Spartan-3E) and Nexys3 and Atlys (Spartan-6) need Xilinx ISE 14.7, discontinued since 2013 and hard to install on the computers and tablets students use today. The author presents Silinx, a free (AGPL-3.0) HDL design platform for teaching that keeps ISE's Project Navigator workflow but runs in a web browser with nothing to install. A module can be written in VHDL or Verilog, in C or C++ through Silinx HLS, which is compatible with AMD's Vitis HLS and passes all 73 of its introductory examples, or drawn as a schematic, an ASM chart, a state diagram, a Boolean function with its Karnaugh map, or a Petri net, from which its HDL is generated. Designs are checked with explanations written for beginners, simulated, run on an emulated board, built, and programmed onto a real board over WebUSB. The implementation flow is open and runs in the browser: Yosys and nextpnr compiled to WebAssembly, with Silinx's own device data, timing model and bitstream writer for the Spartan-3E and Spartan-6, built by black-box observation of ISE's documented tools; Lattice iCE40 boards use the existing open iCE40 tools, and further families are planned. ISE is the oracle: Silinx's bitstreams are byte-identical to those ISE writes for the same placed and routed design, and pass ISE's design rule check. Against ISE at the same optimization goal, the open flow uses 15% fewer slices with 14% shorter critical paths when optimizing for area, and 21% fewer slices with 8% shorter critical paths when optimizing for speed (geometric means over 45 open-source designs built by both flows). The author also describes how the platform was developed with AI coding agents under test-first project rules.
Read the full abstract
Artificial Intelligence
Creative Priming: Decoupling Process from Product to Escape the Creative Paradox
Many of the most consequential advances promised by AI, from curing diseases and inventing new technologies to formulating new scientific theories, are fundamentally creative acts. Yet creativity is uniquely difficult to learn because it faces a creative paradox: learning to repeat past examples of creative success makes those specific examples more likely to recur, inhibiting the very creativity desired. This work argues that creativity is better modeled as an explicit process rather than as a property of individual artifacts, and proposes to learn the underlying generative principles behind the creative process. In particular, to escape the paradox, we introduce creative priming (CP), wherein a trained priming model generates a prompt-specific primer that assists a response model in being creative, explicitly decoupling the creative process from response generation. The priming model learns through GRPO, rewarding primers that lead to responses that are novel, high-quality, and diverse, thereby reinforcing the process that yields creative responses without reinforcing any specific response.
Read the full abstract
Distributed Computing
Why Kubernetes Cannot Contain a Full Lazy-Pulling Cache: Two Impossibility Results and a Bounded, Honest Design for Model Serving
Lazy container-image pulling makes cold starts of autoscaled model-serving fleets on Kubernetes nearly independent of model size, but it moves the model's bytes into a node-level cache that the platform neither bounds nor accounts for. This article establishes, by experiment, two impossibility results for that cache. A monitoring-side remedy cannot catch its failure: a sentinel readiness probe stays green while 272 of 280 files are unreadable. A platform-side remedy cannot contain it: kubelet, given full disk-pressure visibility, frees 0 bytes by image garbage collection and falls back to evicting pods (26 across the campaign, the reading pod among them), because its reclamation actuator is keyed to an accounting that does not know the cache exists. The article then eliminates the failure class inside the component that owns the bytes, with no change to kubelet or the container runtime interface, by two design rules: a byte budget whose pressure signal is keyed to filesystem truth, so that a never-blocking, lossy accounting still enforces a hard bound (held against about 301,000 dropped updates), and pass-through semantics under which a failed cache write never becomes a failed read, with the coupling traced to a single io.MultiWriter. Across six pre-specified campaigns on AWS, upstream stargz-snapshotter fails 273 of 280 model files at 90% cache pre-fill; the proposed system reads all 150.3 GB with zero errors at +18.5% completion cost and holds the budget under a live read, co-tenancy, and a three-node fleet. A source audit shows the missing bound generalizes to SOCI and Nydus. The implementation is open source and has been proposed upstream.
Read the full abstract
Computer Vision
Dense Flow from Adaptive Correspondence
Optical flow requires reliable matching from sufficient visual context and precise motion at every pixel. Existing methods such as WAFT match and estimate flow on the same coarse grids, then recover dense flow through learned upsampling, without explicitly linking visual support for matching to dense flow reconstruction. Our key insight is that dense pixel-level flow does not require equally dense matching. We decouple correspondence scale from flow resolution by adapting the size and shape of spatial supports to image structure for reliable matching, while reusing these supports for precise dense flow reconstruction. We realize this in Flow from Adaptive Token Estimation (FATE), whose token supports expand within coherent regions and contract near motion boundaries. Shared pixel-to-token associations pool features for matching and reconstruct pixel flow from token motions. Flow supervision jointly trains pixel grouping, token matching, and flow estimation without grouping labels, allowing FATE to anticipate which pixels are likely to move together from appearance alone. Trained on FlyingChairs/FlyingThings3D, FATE achieves the highest zero-shot accuracy on Sintel Final and KITTI, a critical setting given scarce real-world flow labels. For visible pixels, FATE improves accuracy within coherent regions, and more near motion boundaries. Its largest gains occur where direct matches are unavailable: Occluded or out-of-frame pixels account for roughly 10% of valid pixels but 50% of WAFT's total error. Experiments show that adaptive correspondence through pixel-to-token grouping enables reliable matching and precise pixel-level flow, improving generalization to unseen real scenes.
Read the full abstract
Functional Analysis
Fredholm weighted composition operators on the Lipschitz space of a tree
Let $\mathcal{L}$ be the Lipschitz space of a tree. In this paper, we establish a complete characterization of the weighted composition operator $W_{\psi,\varphi}$ to be Fredholm on $\mathcal{L}$. Applying our characterization, we describe the connection between the Fredholmness of $W_{\psi,\varphi}$ and that of the composition operator $C_\varphi$.
Read the full abstract
Analysis of PDEs
The Alt-Phillips Multiple Membrane Free Boundary Problem
Motivated by diffusion–reaction processes in porous catalysts, we study an Alt–Phillips model for ordered membranes with weighted gap interactions. The model connects adhesive harmonic junctions \cite{DSS24} and forced multiple membranes \cite{SY19} through the exponent $0<\gamma<1$. For arbitrary positive weights, we prove optimal $C^{1,\beta-1}$ regularity, where $\beta=2/(2-\gamma)$, and obtain nontrivial $\beta$-homogeneous blowup limits through Weiss almost monotonicity. Under strict discrete concavity of the weights, we establish a gluing principle, minimality of flat single-gap cones, and $C^{1,\alpha}$ free boundary regularity at points admitting such blowups. We also classify all one-dimensional vertically critical cones. For three membranes, we identify cone families that are non-minimizing for small $\gamma$ and minimizing near $\gamma=1$.
Read the full abstract
Computer Vision
Correctness Metric Matters: NWD-Based Calibration Loss for Tiny Person Detection in UAV Infrared Imagery
Detection calibration aligns predicted confidence with empirical correctness. Existing calibration losses inherit their correctness definition from box regression, specifically IoU against a hard threshold. We show that this target is geometrically ill-posed for tiny objects in UAV infrared imagery, where persons occupy less than 0.1% of the image. The same pixel error is penalized more heavily as the target shrinks, so binary prediction calibration (BPC) actively suppresses confidences that should be high. On HIT-UAV, BPC with an IoU target degrades image-level localization-aware ECE (LaECE) from 0.0689 to 0.0761, 0.0649 to 0.0733, and 0.0653 to 0.0682 across Low, Mid, and High signal-to-clutter ratio (SCR) groups. We replace the IoU target with a continuous, scale-invariant normalized Wasserstein distance (NWD). The NWD target reduces LaECE\(_{\mathrm{NWD}}\) relative to BPC+IoU on all three groups (\(p<10^{-7}\)) with mAP neutral (0.9356 vs.\ 0.9347), and an independent center-offset anchor confirms that the gain is a genuine geometric correction. The improvement holds across a broad plateau of calibration weights. The NWD formulation requires no inference-time mapping layer. If a validation-fitted isotonic map is applied on top of it, calibration is lowered further, and the two corrections act on different components of the same error. The metric-level remedy has a measurable boundary. Target-background thermal pairing in the training distribution and detection-level aggregation under per-image detection-count heterogeneity lie outside it. Both are quantified independently of the NWD target.
Read the full abstract
Artificial Intelligence
A Bifurcation-Theoretic Framework for the Controllability of Emergent Intelligent Systems
Real-world intelligent systems must reconcile two demands: the adaptability conferred by emergence and the steerability required for accountable deployment. We show that these demands are governed by two distinct bifurcations on the same object and that, consequently, increased capability does not imply increased controllability. Emergence is a bifurcation of the capability variable $S$, with criterion $\Gamma:=\mathcal F'(0)>1$ equivalent to the existence of a unique positive stable fixed point of a self-consistency equation, following solely from the concavity of $\mathcal F$. Controllability is a bifurcation of the permission variable $c$, whose dynamics include an erosion term and exhibit a saddle-node fold with hysteresis as the gate criterion $\rho^\ast$ crosses unity. Three quantitative results follow: the critical exponent distinguishes families rather than classifying the transition; the gating-cost bound $K(M)\propto\log(1+M)$ grows with scope, not capability; and the fold threshold is $\gamma_c\approx4\mu\kappa$. Calibrated in both directions, the criteria reproduce the two-dimensional Ising transition, admit two true-feedback order parameters, and reject nine candidate quantities, supporting the conclusion that the order parameter is geometric rather than capability-related. Since permission cannot be endogenized and no deployed system has been measured to satisfy $\Gamma>1$ and $\rho^\ast>1$ simultaneously, increased capability does not imply controllability.
Read the full abstract
Artificial Intelligence
Can the Remaining Answers Change the Prediction? Worst-Case Stability Checks for Personalized Rating Models
Missing ratings can overturn a personalized model's current prediction. For completed models with additive response scores, exact worst case bounds check all missing assignments without changing the current predictor. We evaluate this audit retrospectively in two self-triage cohorts, 108 clinicians rating three simulated psychiatric interviews, and 126 counseling demonstrations annotated by ten experts. Under shared random half histories, Fisher adaptation based on observed ratings reduces log loss relative to response frequency by 0.1071 nats in the clinician audit (95% CI [0.0867,0.1261]) and 0.0223 in counseling (95% CI[0.0121,0.0321]). The counseling test excludes target experts and source videos from training. Certification covers 5.4% and 10.5% of targets; fixed width acquisition raises coverage to 13.7% and 28.6% at the same budgets. With weights and responses held fixed, the observed-rating update also has lower loss than an affine partial update while certifying exactly the same evaluated targets. The certificate establishes agreement with the completed model, not correctness or stability at intermediate updates. These rating tasks test the method; clinical usefulness remains untested.
Read the full abstract
Robotics
Particle Filtering with Generative Trajectory Models for State Estimation and Control
Generative models can support both state estimation and control in robotics. For control, a common approach is to use the model as a policy that directly samples actions conditioned on observation histories. However, models that jointly represent states, actions, and observations can also predict how the robot state and its environment evolve. We investigate whether these predictions can improve state estimation and control through a change in inference procedure, without modifying the pretrained model. To this end, we use particle filtering to maintain competing hypotheses about state histories. A single pretrained diffusion model generates these hypotheses along with their predicted observations and associated action proposals. As new measurements arrive, particles are reweighted according to how well their predicted observations match the measurements. While particle filtering conventionally uses these weights only for state estimation, we also use them to guide action selection, favoring proposals associated with hypotheses that better explain the observations. Simulation experiments in partially observed navigation and vision-based manipulation demonstrate improvements in both state estimation and task performance compared to action generation without filtering, using the same pretrained model and no additional test-time information.
Read the full abstract
High Energy Physics (Theory)
The tertiary invariant and discrete Green-Schwarz couplings in heterotic string theory
The tertiary invariant is a recently introduced pairing between two-dimensional $\mathcal{N}=(0,1)$ supersymmetric quantum field theories, motivated by topological modular forms. We evaluate it directly from its definition for the sigma models on $\mathrm{Sp}(2)$ and $\mathrm{SU}(3)$ with unit $H$-flux, paired with the internal theories of the non-supersymmetric $\mathfrak{so}(16)\oplus\mathfrak{so}(16)$ heterotic string and of the near-horizon region of the non-supersymmetric heterotic 6-brane, respectively. Regarding the target spaces as $S^3$ bundles over spheres, we fill in every fiber with a four-disk carrying a gauge instanton and reduce the invariant to the index of a Dirac operator on the base. We find nonzero values of order three and two, which determine the Green-Schwarz couplings of the two theories on $\mathrm{Sp}(2)$ and $\mathrm{SU}(3)$, respectively. The former, a $\mathbb{Z}_{3}$-valued discrete topological term of the $\mathfrak{so}(16)\oplus\mathfrak{so}(16)$ theory, was previously studied through anomaly inflow to NS5-branes. Our computation, which uses only the worldsheet theories, agrees exactly with that result and thus confirms the anomaly inflow.
Read the full abstract
Systems and Control
Scarcity Pricing for State-of-Charge Management in Real-Time Markets Under Intra-Day Uncertainty
The growth in renewable resources will increase net load variability and forecast error. Energy storage resources can provide flexibility when ramp capability is limited and contribute to load leveling over the day. The challenge for system operators is how to balance these two functions in real time, when updated net load forecasts are available for a limited lookahead horizon. U.S. electricity markets are seeking ways to manage state-of-charge (SoC) within their energy markets to more efficiently utilize these resources. We propose a method to manage energy storage SoC in a real-time market by extending Hogan's (2013) scarcity pricing framework. A day-ahead SoC schedule is modified using existing multi-interval real-time unit commitment engines as net load forecasts change. To address 30-minute-ahead forecast errors, an updated 5-minute SoC target and a dynamically updated shortage price for violating the constraint are provided to each economic dispatch. We simulate PJM's energy markets for 365 days and demonstrate reliability improvements relative to a strategy that enforces the day-ahead schedule. We show cost reductions and improved price formation relative to a strategy that updates target quantities but uses an administratively chosen shortage penalty.
Read the full abstract
Optics
Wavelength-resolved polarization characterization of supercontinuum vector beams
The polarization state of a supercontinuum (SC) vector beam provides a reference for quantitative polarimetry. We characterize an azimuthally polarized SC vector beam generated with an S-waveplate and azimuthal polarizers by wavelength-resolved Stokes imaging. Representative maps show an approximately tangential azimuthal angle and a small ellipticity parameter over much of the beam, together with the degree of polarization. The {\pm} 90{\deg} azimuthal angle boundary shifts by about 21{\deg} between 450 and 766 nm. A linear fit gives 21{\deg} {\pm} 7{\deg}, where the uncertainty is derived from the slope confidence interval and does not include uncharacterized systematic errors. Jones-matrix calculations examine candidate imperfections under specified assumptions. Attributing the measured boundary shift to this mechanism requires an intensity extinction ratio of about 7 or less at 766 nm if the offset at 450 nm is negligible and the shift corresponds to a uniform azimuthal angle change. The origin remains unresolved. The results motivate wavelength-dependent characterization and calibration of this probe for quantitative polarimetry.
Read the full abstract
Differential Geometry
Isoperimetric comparison for surfaces with nonnegative curvature and uniformly convex boundary
We establish a sharp relative isoperimetric comparison for compact surfaces with nonnegative Gaussian curvature and smooth nonempty boundary whose geodesic curvature is at least 1. We prove that the normalized relative isoperimetric profile for such a surface is bounded below by that of the Euclidean unit disk, with equality holding at some area fraction in (0,1) if and only if the surface is isometric to the unit disk. The proof uses a differential comparison argument for the isoperimetric profile, based on a suitable non-constant normal variation of isoperimetric curves inspired by the model geometry.
Read the full abstract
Artificial Intelligence
State Media at Retrieval Time: A Large-Scale Audit of Russian State-Aligned Sources in AI Chatbot Responses
The Pravda Network is a Russian state-aligned media aggregator, with over 110 websites and more than 15 million articles published since 2023. Given its prolific and repetitive output, emerging scholarship argues that the Pravda Network was built to influence AI systems. We conduct a large-scale audit of leading AI chatbots to understand: (1) the retrieval rates of Pravda Network articles, (2) the conditions of retrieval, and (3) whether chatbot responses affirm or caveat this content. We build a 10,000 prompt corpus from sampled Pravda Network headlines varied along two dimensions (headline rewording and user skepticism) and query 11 AI chatbots via the web interface (n=104,997 completed responses). Overall, we find susceptibility to Pravda Network retrieval to be infrequent—appearing in 7.3% of responses—and highly platform dependent—with three chatbots responsible for 94% of retrievals (You.com, Claude, and Copilot). Among the frontier models, Claude is an outlier: ChatGPT, Gemini, and Meta AI retrieved no Pravda Network articles, likely indicative of a domain-level filtering policy. All 11 chatbots, however, retrieved other Russian state-aligned media to some degree (median=13.56% of responses). When Pravda Network articles are surfaced, caveat rates vary widely by platform (from 5.3% to 76.2%), with You.com providing the fewest caveats. We also identify the tactic of headline washing: where 9.92% of sampled Pravda Network headlines attribute their reporting to other, often more credible, media outlets, while the articles themselves are republished from Russian state-aligned media. We discuss headline washing's potential to launder state media as independent reporting. With this audit, we provide a systematic analysis of AI chatbots' susceptibility to Russia's latest attempt to manipulate the information environment.
Read the full abstract
Machine Learning
The Dark Matter of Clinical Data: An Identification Limit on Disentangling Physiology from Care Process in electronic health records
Representations learned from electronic health record (EHR) data encode patient physiology together with the clinical process that decides what is measured, when, and for whom. We formalize the data-generating process with latent physiology, ordering intent, laboratory values that exist only when measured, a learned representation, and clinical environments, and analyze two strategies for undoing this entanglement. First, no division of the information carried by laboratory "value" and "ordering" channels is identified from a single dataset: when ordering depends on the analyte, its distribution among untested patients is not recoverable, and a sentinel code for the undefined recorded value makes it carry all the entropy of the ordering indicator, fixing partial information decomposition (PID) for every redundancy measure satisfying the Williams-Beer axioms and invariant under injective re-encoding. Second, cross-site invariance of representation-conditional outcomes is a falsifiable implication of biological invariance taken jointly with exclusion and representation-sufficiency conditions, but not an identification strategy: we construct observationally equivalent models that disagree about biological invariance, whether the invariance test fails or passes. For binary physiology this happens exactly when case mix differs, the representation is lossy, and physiology affects the outcome. A worked example yields a 9.6-percentage-point cross-site gap under an unchanged biological mechanism. Tests performed for every patient remove the ordering gate, and measured tests can restore identification: biological invariance is identified when a test exactly summarizes outcome-relevant physiology, or, for binary physiology, when two tests with independent errors are available at sites with different case mix. Cross-site analyses should be reported as tests of a stated joint hypothesis.
Read the full abstract
Artificial Intelligence
The Value of Not Knowing: How Fresh Eyes through Naive LLMs Fuel Creativity
Many major scientific advances come from approaching a problem with a new or novel perspective. While large language models (LLMs) with internet-scale knowledge seem poised to take such perspectives, their consistent generation remains elusive. Instead, models are drawn to dominant views and preexisting, in-field bias. To explore the potential of fresh eyes to increase creativity, this paper proposes naive LLMs, models that are made intentionally ignorant of a specific topic. Through deprivation in training or intentional forgetting, naive models are unburdened by existing dogma, allowing them to be free to make new associations and conjectures. These naive models can then be paired with a stronger mentor model that can refine their unique perspectives into concrete proposals. For the 1B and 7B models tested in this paper, naive models consistently produce seed ideas that are more novel and useful than baseline counterparts, and mentor refinement improves the ideas even further. Thus, contrary to the prevailing tradition of scaling up training data, we demonstrate a case where less information is actually more powerful, in particular when the aim is to discover genuinely outside-the-box solutions.
Read the full abstract
Computation and Language
Explaining and interrupting collective misalignment in language model systems
In collectives, language model agents can exchange individually innocuous messages that together support a policy violation. Understanding and preventing collective misalignment requires answering the question of which messages sufficiently influenced and necessarily enabled the violation. Collective Dependence Analysis (CDA) operationalises necessity and sufficiency to explain decisions and find the fewest removals interrupting support within tested combinations. Tests covered constructed violations, unscripted agent conversations, and recorded interactions, with selected message removals reducing violations from 93% to 6% in controlled cases. Using O(N) model evaluations for N messages, CDA achieves the same violation reduction with half the deletions of a frontier LLM judge given the transcript and asked for all supporting messages in eligible operational cases. CDA also proposes the mechanism through which the collective misalignment emerged. Nonetheless, complete explanations remain difficult on unfamiliar tasks and some removals can inadvertently disrupt correct decisions. CDA unifies explanations of collective misalignment with experimentally tested interventions.
Read the full abstract
High Energy Physics (Experiment)
Measurement of the branching fractions for the decays $\chi_{cJ}\to \phi \omega \eta$ and $\chi_{cJ}\to \phi \omega \eta^{\prime}$
Using a sample of $(2712.4 \pm 14.3) \times 10^{6}$ $\psi(3686)$ events collected with the BESIII detector at the BEPCII collider, we report the first observation and measurement of the branching fractions for the decays $\chi_{cJ}\to \phi \omega \eta$ and $\chi_{cJ} \to \phi \omega \eta^{\prime}$ ($J = 0,1,2$) with statistical significance greater than $5\sigma$. The branching fractions are determined to be $\mathcal B(\chi_{c0} \to \phi \omega \eta) = (5.43 \pm 0.23 \pm 0.52) \times 10^{-4}$, $\mathcal B(\chi_{c1} \to \phi \omega \eta) = (2.51 \pm 0.14 \pm 0.24) \times 10^{-4}$ and $\mathcal B(\chi_{c2} \to \phi \omega \eta) = (3.03 \pm 0.15 \pm 0.28) \times 10^{-4}$ for the $\eta$ mode, and $\mathcal B(\chi_{c0} \to \phi \omega \eta^{\prime}) = (1.23 \pm 0.23 \pm 0.15) \times 10^{-4}$, $\mathcal B(\chi_{c1} \to \phi \omega \eta^{\prime}) = (1.09 \pm 0.17 \pm 0.13) \times 10^{-4}$ and $\mathcal B(\chi_{c2} \to \phi \omega \eta^{\prime}) = (0.98 \pm 0.17 \pm 0.11) \times 10^{-4}$ for the $\eta^{\prime}$ mode, where the first uncertainties are statistical and the second systematic.
Read the full abstract
Differential Geometry
The topology of end-periodic smoothing operators
We develop an algebra of smoothing operators adapted to the geometry of end-periodic manifolds. We extend the renormalized trace and trace defect formula of Mrowka-Ruberman-Saveliev to this algebra, and establish a Fredholm criterion and winding-number index formula for smoothing perturbations of the identity. Our main result computes the homotopy groups of the associated end-periodic smoothing group. This provides a foundation for future work on relative index theory for families of end-periodic differential operators.