Papers

A group does not usually consult all its members before every decision. This paper develops a quantitative framework for measuring how widely members are consulted in decision making within a group. Two kinds of records are considered: the number of members consulted for each decision and the number of decisions in which each member is consulted. From these records, two participation indices are defined, and their equality is proved analytically. A third index measures the fraction of decisions for which at least a specified percentage of members are consulted. The distribution of consultation levels is described using a discrete maximum-entropy model from statistical physics. Tail probabilities are used to obtain bounds for the main participation index. This study examines aggregate survey data from two groups. Their observed characteristics are broadly consistent with the patterns found across the full survey. Using binomial and Poisson calculations, the future probability that all members will be consulted for a decision is examined. Inequality in consultation among members is examined through Lorenz curves, the Gini coefficient and a normalised Shannon participation measure. The member-decision structure is represented as a bipartite network. Controlled simulations show that the same overall level of consultation can have different distributions among the members. The framework measures both the overall breadth of consultation and how consultation is distributed among the members. It therefore provides a quantitative sociophysical description of one measurable aspect of democracy in decision making within a group.

While Vision-Language-Action models have enabled impressive cross-task transfer in zero-shot settings, adaptation methods are often necessary to achieve consistent performance on novel tasks. However, even though a wealth of methods for adaptation have been proposed, it is unclear how to choose the best method based on available trajectories. We close this gap by studying the impact of the adaptation method under a variety of common data scenarios: on- and off-policy data, offline vs online adaptation, and whether the data contains optimal or suboptimal trajectories. Our evaluation spans simulated tasks in the LIBERO suite and real tasks on a Franka platform. In the offline data regime, critic-based steering approaches outperform supervised finetuning when failure data is available. However, offline RL techniques for training critic functions fail to outperform simpler Monte-Carlo-based approaches. Finally, in the online setting, critic-based steering is significantly more sample efficient compared to policy-gradient-based finetuning methods, but has a lower performance ceiling than full finetuning.

Graph Neural Networks on large graphs commonly rely on in-memory graph storage with ad hoc data processing pipelines for both training and inference. Graph databases provide scalable storage and efficient traversal for graphs that can evolve over time, but their applicability in GNN training and inference are rarely studied. We show that graph DBs can be used in GNN training and inference directly, executing sampling, feature retrieval, and even the full forward pass. We show different GNN training pipelines can perform on-the-fly sampling and retrieval in graph DBs using Cypher. We further provide optimisations that speed up the training process by \~2.5x while maintaining the same accuracy. We illustrate that fully in-DB inference is possible, independent of how the GNNs are trained. Our inference setup can be >10x faster than cold loading the graph into RAM with no extra memory overhead on real-world graphs. Our work lays the foundation for in-graph-DB GNN training and inference.

Learning a collective policy for state-based Cartesian manipulation across diverse robotic embodiments presents a fundamental challenge due to embodiment-dependent transition dynamics. Collective learning aggregates multi-robot experience to discover shared task structure, yet naive aggregation can average incompatible action modes when different embodiments realize the same Cartesian command through different transition behavior. This leads to strategy ambiguity and performance collapse on dynamically outlying embodiments. To address this, we propose PACE, a two-stage collective learning framework that separates shared task-space intent from embodiment-dependent transition dynamics. The first stage employs a representation layer pre-trained on offline trajectories to extract task-level context from collective data. The second stage introduces a decision layer equipped with a predictive dynamics adapter. By forcing the latent state to predict future transitions, this adapter acts as a latent regularizer that grounds the policy in the transition behavior of the executing robot while preserving real-time inference without online planning. Evaluations across simulated embodiments demonstrate that our method prevents performance collapse on distinct robots. It achieves over 93% overall success and 89% on dynamically out-of-distribution zero-shot simulated robots, outperforming the Transformer behavior cloning (82%) and diffusion policy baselines (66%). Physical KUKA deployment further validates the feasibility of the full state-based pipeline under zero-shot sim-to-real transfer.

Tracing my scientific path from many-body physics in semiconductor nanostructures to topological quantum matter, this Habilitation à diriger des recherches (HDR) memoir presents only part of my research since defending two PhD theses in Toulouse and Nizhny Novgorod in 2011. Focusing on pseudo-relativistic fermions and topological states in zinc-blende semiconductors and their heterostructures, it highlights my independent research program: identifying physical problems, developing predictive theories, and initiating new directions with experimentalists. The first part addresses Kane fermions in bulk HgCdTe: temperature-driven band inversion, domain-wall states, the collapse of Landau levels in crossed electric and magnetic fields, and disorder-induced topological transitions. The second examines pressure-, temperature-, and disorder-driven phase transitions in single and double HgTe quantum wells (QWs), including bilayer-graphene-like semimetals and a two-dimensional higher-order topological insulator (2D HOTI) phase beyond the minimal BHZ model. The third develops three-layer InAs/Ga(In)Sb QWs as a III-V platform for massless Dirac fermions, large-gap quantum spin Hall insulators, 2D HOTI phases, and semimetals with quadratic band touching, and discusses InAs$_{x}$Sb$_{1-x}$/InAs$_{y}$Sb$_{1-y}$ heterostructures as a promising alternative. Several of these predictions are supported by experimental observations, including quantized helical edge transport up to 60 K and the unconventional quantum Hall effect. Defended on March 25, 2026, at the University of Montpellier. Jury: Michel GOIRAN (chair and rapporteur, LNCMI-CNRS, Toulouse); Alexey KAVOKIN (rapporteur, Westlake University, Hangzhou); Robson FERREIRA (rapporteur, LPENS-CNRS, Paris); Alexei BARANOV (member, IES-CNRS, Montpellier); Benoit JOUAULT and Frédéric TEPPE (members, L2C-CNRS, Montpellier).

Tokenization is an essential preprocessing step in almost all large language models (LLMs). One of the most popular frameworks is BPE tokenization, which consists of special token segmentation, pretokenization, and byte-pair encoding (BPE). However, existing algorithms are sequential, with each stage only processing complete output from the previous stage. In order to reduce latency and memory footprints, recent works (Jiang and Gong, ICML 2026) (Li, Yang and Mamouras, ASPLOS 2026) (Mamouras, Li and Yang, PLDI 2026) have initiated the study of online algorithms for subroutines such as pretokenization and BPE with a goal of streaming input to output as the input arrives online. However, the search for efficient online tokenization remains open. In this work, we formalize and resolve online tokenization. We introduce the notion of delay for online tokenization which characterizes the extent to which the algorithm output lags behind the (possibly computationally intractable) optimum. We give a near-linear time algorithm with optimal delay (up to constant factors). We complement this algorithm with a conditional lower bound: unless $\mathbf{P} = \mathbf{NP}$, there is no polynomial time algorithm that beats the delay of our algorithm by more than a constant factor. We implement a practical variant for tokenizers from OpenAI's tiktoken and HuggingFace's tokenizers. We obtain a $7.9\times$-$34.8\times$ latency advantage (measured as time to first token) over tiktoken (as well as the recent gigatoken tokenizer) and our implementation maintains approximately constant memory usage independent of input size. Our implementation also achieves a $2.4\times$-$7.0\times$ improvement in throughput over tiktoken and $15.8\times$-$38.1\times$ over tokenizers.

How well can a language model turn a large, incomplete and partly contradictory body of documents into a plan when no single correct answer exists? We introduce Mission Analysis (MA)-Bench, built on mission analysis, the first step of military operational planning, in which a staff reads the full situation and states the problem before anyone proposes a solution. Its 10 expert-built scenarios pair synthetic packets of ten to twelve documents with 1,257 expert-written pass/fail requirements, each containing information on their source, question type, and importance level. A reasoning-model judge grades each brief against the requirements without seeing the documents; it grades like a fifth credentialed planner, leaving a four-planner panel's agreement unchanged (Fleiss $\kappa$ 0.42 to 0.41), and a judge from a second provider reproduces the model ranking (Spearman 0.94). Across twenty frontier and open-weights models the best gets 62% of the rubric items right. Models are strongest on facts stated in the documents (58%) and weakest on judgments that must be inferred (33%), and six failures in ten are omissions rather than errors. MA-Bench grades the determinations, not the prose. We release the packets, rubrics, prompts, scorer and every graded brief with item-level verdicts.

Decarbonizing cellular networks requires coordinating communication-resource and battery operation with time-varying electricity prices and carbon intensity. We jointly optimize base station and battery operation under a carbon-adjusted electricity price and traffic-demand constraints, and solve the mixed-integer nonconvex problem through a two-stage decomposition exact under a minimum-power rule. Using California price and carbon-intensity data, we show that carbon-aware scheduling reduces emissions most when low-price and low-carbon periods are misaligned. Above a carbon price of roughly 100-300 USD/ton, battery operation shifts from cost-driven to carbon-driven, cutting emissions by up to 57.8% relative to a cost-only schedule.

We present a graph-based framework for Composite AI systems that integrate multiple AI techniques to address complex problems. We describe an architecture using specialized control nodes to orchestrate diverse models including planning, machine learning, constraint programming, and large language models. The system employs directed acyclic graphs and behavior trees to manage model relationships and execution sequencing. Our methodology encompasses data connectors, computational services, and a containerized deployment framework supporting hierarchical, sequential, and concurrent model composition patterns. Through case studies, we demonstrate how this architecture enables the creation of composite systems that leverage the strengths of multiple AI techniques within a unified operational schema, overcoming the limitations of single-model approaches.

This paper extends the applicability of probabilistic graphical models to non Gaussian data. It establishes a characterization of the pairwise Markov property in undirected probabilistic graphical models in terms of copula densities. Building on this characterization, we derive a necessary condition for the pairwise Markov property in Student copula Markov networks. An Expectation-Conditional Maximization algorithm (ECM) is provided to facilitate the penalized maximum likelihood estimation and graph selection. The performance of the proposed algorithm is evaluated through simulation studies, which demonstrate its effectiveness in estimating model parameters and recovering the underlying graph structure.

Feature distillation is usually expected to add to logit distillation. We find that this expectation does not hold for two common feature losses in compact multilingual automatic speech recognition (ASR), and analyze why through their gradients. The gradient of a projection-based loss is orthogonal to the task gradient, while that of Centered Kernel Alignment (CKA), a similarity-based loss, is weakly aligned with it. CKA lowers the word error rate (WER) on non-English languages, but at the expense of data-rich English. Three loss-level remedies (one trained, two probed) all fall short, and the CKA gradient overlaps with the subspace English uses in the shared encoder by 97.5% to 99.5% (about 81% by chance). Guided by this diagnosis, we propose CAFE (Confined-Adapter Feature distillation), which confines the CKA gradient to adapters that English bypasses, addressing the trade-off in the parameters rather than in the loss. On the Open ASR Leaderboard, CAFE lowers the average WER from 8.55% to 8.02%, beats an adapter-only capacity control at 8.51%, and makes English more stable across checkpoints. At 254M parameters, it trails the 0.6B teacher by 2 WER points and is competitive with much larger models across 48 test sets in 24 languages. We use 564k hours of training data and state-of-the-art models to keep the findings relevant to practice.

Indirect prompt injection uses untrusted retrieved content, tool outputs, or agent messages to override an LLM's intended behavior. We show that defense robustness is channel-conditioned: the same attack succeeds 18% of the time as a follow-up message but 53% when placed in retrieved context. We propose TRACER (Trace-Regularized Adversarial Channel Exposure with Reward-shaping), a channel-aware adversarial training method that samples presentation wrappers and combines an answer-anchored, evidence-gated reward with ATRC, which aligns attacked reasoning with clean evidence when an injection is invalid. On GSM8K with Qwen2.5-3B-Instruct, TRACER reduces RAG-framed ASR from 65.5% to 6.1% while preserving 78-82% clean accuracy. Joint RAG+tool training achieves 13.9% average ASR across framings at 3B scale, while a single epoch at 7B reaches 9.4%. These results demonstrate that robust defenses must explicitly train and evaluate across deployment channels.

For $1\le p < \infty$, $\varphi_p$ is the function of minimal $L^p$ norm among all functions $f$ of exponential type at most $\pi$ for which $f(0)=1$. We start from the ansatz that the positive zeros $\tau_n$ of $\varphi_p$ are close to \(n+1/p-1/2\) and use a weighted convex correction to first place them in an \(\ell^2\)-neighborhood of this sequence. A bootstrap argument then improves this to a {Poincaré asymptotic} expansion of $\tau_n-(n+\frac1p-\frac12)$ in terms of odd powers of $(n+\frac1p-\frac12)^{-1}$. We also get precise information about the Fourier transform $\widehat{\varphi}_p$, a result that in particular implies that for all integers $n\ge 2$, $\widehat{\varphi}_{2n}$ is an integrable function which, when convolved with itself $2n-2$ times, is constant on the interval $[-\pi,\pi]$. In particular, we identify the extremizer in an autoconvolution problem posed by B. Green in 2001.

Maintenance work orders are the largest written record of how industrial equipment fails, yet in most plants they are close to unusable for analysis. Short texts are abbreviated and misspelt, asset identity is buried in free text, work-type labels contradict the work described, and the fields that matter most for reliability (symptom, root cause, action taken) are usually empty. As a result, diagnostic knowledge stays with senior technicians and leaves when they do. We present Plant Memory, a system that applies technical language processing (TLP) to SAP Plant Maintenance (PM) exports and builds an ISO 14224-aligned maintenance ontology in which every extracted statement is a reified fact carrying a confidence score, the evidence it was read from, and a verification state. Extraction is rule-first, with an optional large language model (LLM) acting as a second reader whose disagreements raise a fact's review priority rather than overwrite it. Technicians verify facts through a review queue, answer targeted knowledge-gap questions where failures repeat without a recorded cause, and approve or reject ontology drift proposals when new terms appear. The resulting graph is served to people through a web application and to AI agents through a Model Context Protocol (MCP) server, so both query the same verified knowledge and every answer cites the work orders behind it. On a synthetic 12-month export of 400 work orders from a parboiled-rice plant, rule-based extraction reaches 96.2% component accuracy, 88.9% failure-mode accuracy and 100% root-cause precision at 62% coverage; on 28 diagnostic questions, ranked root causes from the ontology are correct at top-1 in 82.1% of cases against 53.6% for the raw SAP root-cause column.

We apply large $N$ expansion techniques to derive kinetic equations for strongly coupled, far-from-equilibrium field theories with $O(N)$ symmetry. For large $N$ vector models, the quasi-classical approximation holds. Incorporating arbitrary momentum-dependent interactions, we show the dynamics obey the standard kinetic equation, with the scattering amplitude given by the retarded propagator of the auxiliary Hubbard-Stratonovich field. This has broad applications to systems like spin wave and Langmuir ion-acoustic turbulence. For large $N$ random coupling (SYK-like) models in $d$ dimensions, leading-order melon diagrams induce order-one propagator renormalization and finite spectral width. The corresponding kinetic equation, unlike those for standard weakly interacting systems, does not have a local-in-time collision integral. We find evidence for a stationary, scale-invariant infrared solution, revealing a potentially novel strong turbulence scaling exponent.

Training from offline data has allowed for substantial progress in domains such as robotics, leading to general-purpose policies that can be easily applied zero-shot or efficiently finetuned for downstream tasks. However, these policies can still have poor generalization, both due to the choice of modeling objective and from learning from a static dataset. In this work, we focus on the challenging task of zero-shot goal generalization, where a policy is evaluated on unseen tasks that require combining its existing knowledge (compositional generalization). An avenue for improving a policy's generalization is by generating new experience with world models; however, such generation has proven difficult for longer horizons. Thus, to alleviate this issue, we propose TD-Aug, which samples from a geometric horizon model and allows for directly imagining novel outcomes that can be achieved by composing existing knowledge. We demonstrate that training on these future outcomes as goals for goal-conditioned BC policies significantly improves generalization in stitching-based OGBench tasks.

Vision-language models (VLMs) are increasingly deployed in multi-agent systems to automate complex computer-use workflows, yet existing benchmarks mostly assume consistent, cooperative environments. In contrast, real-world collaboration rarely runs on consistent information: peer or sub-agents frequently report outdated states, issue contradictory instructions, or compete over shared permissions and files, forcing agents to check others' outputs and verify ground truth in the environment. Evaluating such conflict resolution remains challenging: manual authoring cannot scale across complex systems, while unconstrained LLM generation produces narrative drift, answer leakage, and unsatisfiable rubrics. We introduce CLASHarness (Conflict-Loaded Agent Scenario Harness), a controllable framework for generating verifiable multi-agent conflicts in computer use, with a benchmark of 600 multi-round tasks. By combining hierarchical scenario planning with an eight-step verified micro-pipeline, CLASHarness pre-commits decisive visual facts before rendering and compiles hybrid code and semantic graders directly from executed reference trajectories, guaranteeing every task is solvable by construction. Evaluating 15 frontier models on CLASHarness reveals three core bottlenecks. First, when sub-agent outputs conflict with the environment, weaker models blindly follow misleading information, whereas stronger models inspect visual evidence to verify ground truth. Second, resolving conflicts is much harder than finding them: while agents can spot inconsistencies, actively fixing system state drops task success significantly. Third, coordination protocols remain a universal bottleneck: even frontier models fail over 58% of closing actions, often forgetting to release shared locks or hand off tasks. CLASHarness provides a scalable, verifiable foundation for building reliable multi-agent systems.

Uniform frame sampling for vision-language model (VLM)-based video analysis often floods multimodal backbones with redundant content, while existing keyframe selection methods are typically optimized for human viewing rather than downstream extraction accuracy. We propose AdaFrame, a content-adaptive preprocessing cascade that reduces visual redundancy without sacrificing extraction fidelity. Three progressively finer filters, perceptual hashing, learned perceptual image patch similarity (LPIPS), and CLIP-based semantic clustering, efficiently eliminate redundant frames before costly cross-modal inference. Per-video frame budgets are determined using Intrinsic Semantic Dimensionality (ISD), which estimates each video's semantic complexity through singular value decomposition of its frame embeddings. Experiments on 500 video advertisements from the Pitt Ads Dataset show that AdaFrame reduces the frame count by 48 percent and VLM inference cost by 75 percent compared with uniform sampling, while achieving 86.0 percent topic accuracy, an improvement of 1.6 percentage points, and 79.4 percent sentiment accuracy, an improvement of 7.5 percentage points.

We analyzed a sample of 494 galaxies from the Sloan Digital Sky Survey (SDSS) I-IV that have unusual properties. They display broadened Mg II $\lambda$2798 line emission, as would be expected of Type 1 active galactic nuclei (AGN). However, they have weak or absent broad Balmer lines though broadened Balmer lines often accompany broadened ultraviolet lines in Type 1 AGN spectra. We performed a spectral decomposition of these "broad Mg II galaxies” to assess whether they can be considered fading AGN. The Balmer lines are more responsive to changes in the AGN continuum than the Mg II line. As the AGN power source fades, broad H$\beta$ and then H$\alpha$ will weaken and eventually disappear, leaving broad Mg II as a signpost of a more active past. In support of this interpretation, we found that broad Mg II galaxies have weaker AGN luminosities and lower accretion rates (parameterized by the Eddington ratio) than typical Type 1 quasars from SDSS. The Eddington ratio of broad Mg II galaxies are similar to dim-state changing-look AGN, with an average value $\lesssim$1%, which is the threshold at which the inner region of a geometrically thin accretion disk is thought to evaporate and be replaced by a geometrically thick radiatively inefficient accretion flow in stellar-mass black holes. Broad Mg II galaxies may represent the penultimate stage in a changing-look AGN sequence prior to the complete disappearance of all broad lines.

Flat-panel loudspeakers are typically driven by dynamic exciters that employ a voice coil coupled to a reaction mass to exert force on the panel. Unlike conventional loudspeakers, where the cone is relatively light weight, heavy panels may require significant exertion from the driver to achieve sound pressure levels required of consumer grade loudspeakers. Overexertion can cause harmonic distortion and reduced sound quality, especially in drivers where the voice coil is not centered in the magnetic field. There are many different exciters available commercially with different power ratings and price points. The frequency response, linearity, total-harmonic distortion (THD), intermodulation distortion (IMD), and force vs. displacement were measured for three types of exciters to determine the consistency of manufacturing and use in panel loudspeakers. These exciters were found to have high variability in all metrics among units of the same type. Distortion metrics such as IMD may be reduced by running impedance sweeps to match exciters with similar response characteristics in multi-actuator panels.

We study the tree-level fermion-photon effective vertex for on-shell particles in a relativistic environment undergoing rigid rotation at zero temperature. The rotational background breaks translational invariance in the transverse plane and introduces a preferred direction, leading to a distinct decomposition of the vertex into longitudinal and transverse structures with respect to the rotation axis. Using fermion and photon modes constructed in the rotating frame, we derive the corresponding selection rules and show that the energy and the longitudinal component of the momentum are conserved, whereas no independent transverse momentum-conservation condition arises. Instead, the transverse dependence is governed by radial integrals involving three Bessel functions, whose support constrains the magnitudes of the transverse momentum projections. The vertex can be organized in a circular basis for the transverse components and a Cartesian basis along the rotation axis, and expressed in terms of six Dirac structures. Their coefficients define 24 form-factor components associated with the photon-mode components, physical photon states, and spin-dependent structures induced by the rotating background. We also verify current conservation and gauge invariance of the resulting tree-level matrix element under the boundary conditions considered. These results provide the tree-level structure of the fermion-photon interaction in rotating relativistic matter and establish the basis for the study of radiative corrections in vortical backgrounds.

Generating novel molecules with desired properties is a central challenge in drug discovery. Although foundational work focused on de novo generation, practical use cases require a more complex setting: conditioning on an initial seed molecule, optimizing multiple properties, and operating in data-sparse regimes. We introduce PairIT, a flexible training framework for autoregressive transformers over SMILES sequences that enables seeded generation by learning from pairs of structurally similar molecules. By leveraging matched pairs, the model learns to generate local, chemically meaningful edits that improve a property of the seed molecule. Prompt tokens extend this framework by specifying the target property, desired direction, and approximate magnitude of change. Multiple prompt tokens can be concatenated to enable multi-property optimization at inference time. For properties where pair supervision alone is insufficient, reinforcement learning post-training can further improve optimization, providing a tunable trade-off between proximity to the seed and the optimization objective. PairIT naturally supports pretraining on a large dataset and fine-tuning on a smaller dataset, which is critical for performance in data-sparse regimes. Surprisingly, we observe diminishing returns to the scale of the pretraining dataset. We demonstrate empirically on MoleculeNet and Therapeutics Data Commons datasets that, with appropriate paired-data curation, a standard transformer model of modest size (68M parameters) can achieve robust seeded molecular optimization in low-data regimes.

Deep space missions increasingly require spacecraft capable of autonomously operating in environments with communication delays and bandwidth limitations. A key enabler of such autonomy is the ability to run learning-based algorithms directly onboard. In this work, we present a flight deployment on a 6U CubeSat that demonstrates (1) onboard training and inference of a thermal segmentation model on an NVIDIA Jetson TX2 NX GPU in space and (2) closed-loop attitude control using the segmentation model on a real-time, thermal image feed to track the Moon. In addition to this demonstration, we implement ROS 2 Foxy as a data transport middleware and integrate an open-source file system that provides strong data integrity guarantees and self-healing features. Over the course of approximately one year, we performed 12 experiments on the spacecraft, advancing the state of the art in autonomous spacecraft systems by showing how modern learning-based pipelines, commercially available hardware, and open-source software can operate reliably under the compute and environmental constraints of Low Earth Orbit (LEO).

Structured pruning reduces LLM inference costs by removing model components to produce smaller dense models. However, existing methods can cause reasoning performance to collapse on long-form reasoning tasks such as AIME. We propose LayerTrim, a training-free structured pruning method that identifies redundant decoder layers and selectively trims their MLP features while preserving attention and decoder depth. LayerTrim is the first training-free structured pruning method designed to preserve long-form reasoning without gradients or recovery training. Across three reasoning language models and five benchmarks, LayerTrim outperforms existing structured pruning methods in most evaluated settings, improving average accuracy by up to 21.4 percentage points over RESP. On Qwen3-14B, LayerTrim achieves 1.24$\times$ faster prefill and 19.7% lower peak GPU memory than the dense model, while better preserving reasoning ability. It also achieves the shortest average generation length among evaluated pruned models, while retaining the highest average accuracy. These results demonstrate that LayerTrim improves reasoning efficiency through both structural compression and reduced token generation.

Purpose: Free-breathing abdominal MRI is limited by respiratory motion, which introduces image artifacts and k-space inconsistencies. Existing reconstruction approaches often depend on pre-estimated motion fields or external respiratory surrogate signals. We propose a coupled reconstruction-motion estimation algorithm for self-navigated free-breathing acquisition, in which motion is represented by compact Gaussian primitives and jointly refined with the image estimates during iterative reconstruction. Methods: An iterative reconstruction that alternates between data consistency updates and motion modelling was proposed. Respiratory motion was represented using Gaussian primitives. A Gaussian splatting was integrated into the reconstruction algorithm. The method was evaluated on self-navigated free-breathing Cartesian gradient-echo (GRE) acquisition in four healthy volunteers, including retrospective scan shortening experiments, and was compared with regularized compressed sensing (CS) reconstruction. Results: Respiratory motion states were reconstructed in approximately 1.5 min. Compared with CS, the proposed method improved visual image quality, particularly in coronal liver views, and better preserved anatomical structures in retrospectively shortened acquisitions. These findings provide proof of concept for the proposed method. Conclusion: The proposed coupled reconstruction-motion estimation algorithm improved visual image quality compared with CS alone. Gaussian primitives provided a compact and sufficiently flexible representation of nonrigid respiratory motion with practical runtime, supporting further scanner integration and validation in larger cohorts.

The Chan–Robbins–Yuen polytope $\mathrm{CRY}_{m+1}$ has normalized volume $\prod_{j=1}^{m-1}C_j$, a product of Catalan numbers. Zeilberger proved the formula using the Morris constant-term identity; finding a combinatorial proof remained an open problem. We give such a proof by counting nonnegative integer flows. We add an extra edge from each vertex to a common sink and count flows according to how many of these edges carry positive flow. A double count proves a special case of a recurrence of Morales and Shi and gives these counts in terms of Narayana numbers. Deleting the first vertex then yields the Catalan product. We also give an explicit bijection for the recurrence and prove the corresponding reduction formula posed by Postnikov.

Emerging decision models such as JEV provide typed interfaces that map task instructions and observations directly to choices, judgments, and scores. As these models are increasingly used for automated decision-making, understanding their reliability under adversarial manipulation becomes critical. We introduce DecisionRobust, a benchmark for systematically evaluating the robustness of decision models across textual and visual inputs. Our benchmark covers practical decision-making scenarios and a diverse suite of attacks, including semantics-preserving text perturbations, image perturbations, and prompt injection in both modalities. We curate evaluation sets through external validity checks and cross-model candidate selection. We evaluate 30 model configurations on text tasks and 17 on vision tasks, including 24 and 12 held-out models, respectively, that are excluded from candidate selection. Our experiments reveal several key findings. First, attacks expose vulnerabilities beyond the models used for construction and selection, although optimized image perturbations transfer weakly to many targets even with multiple surrogates. Second, relative model robustness varies across tasks and attack types, and resistance to input perturbations does not necessarily imply resistance to prompt injection. Third, label-consistency checks substantially reduce measured attack success across text and image settings, highlighting the importance of validating apparent failures. The benchmark is available at https://yifan-lan.github.io/DecisionRobust/.

We carefully construct the profinite iterated monodromy group of a rational map as a quotient of the étale fundamental group of the punctured projective line. In doing so, we prove a number of folklore results regarding profinite IMGs; in particular, their topological generation by inertia concentrated in the post-critical orbit, and their relationship with the discrete iterated monodromy group over $\mathbb{C}$.

We prove the uniform convergence of Plateau borders to Plateau singularities in the vanishing-volume limit of a soap film model based on capillarity theory. The results are formulated in the context a new type of local area minimizer that takes into account how a soap film locally separates its surrounding space.

Any AIS transmitter can claim any vessel's identity (MMSI), because nothing in the protocol authenticates it; only the transmitter's analog hardware reveals who is really sending. We treat AIS traffic as streams of bursts keyed by claimed MMSI and accumulate physical-layer evidence with a sequential test. Each burst is scored by its k-nearest-neighbor distance to enrolled fingerprints in the feature space of a classifier whose weights are fixed after training, calibrated into a tail probability, and accumulated by a geometric-entropy-minimization cumulative-sum (GEM-CUSUM) statistic. The same test rejects unknown transmitters when scored against all enrolled vessels, and exposes insider spoofs when scored against the claimed vessel. A threshold refreshed from genuine traffic alone keeps it at a stated false-alarm rate as fingerprints drift. On AIS in-phase/quadrature (I/Q) recordings captured over 13 days at the Istanbul Strait (Bosphorus), the detector rejects unseen vessels at least 10 days after enrollment. Evaluated over 50-burst trials with every vessel weighted equally, it reaches an area under the ROC curve (AUROC) of 0.956 (95% confidence interval 0.913-0.989), where the best score computed from one burst at a time reaches 0.899. A threshold refreshed from the first quarter of the enrolled vessels' traffic holds 5.3% false alarms (3.0-9.2% across five independently trained encoders) at 88.8% detection. Scored against the claimed MMSI, enrolled transmitters spoofing another vessel are rejected at 0.991, a threat that open-set tests cannot see (AUROC 0.500, chance level). The lightweight Transformer GLFormer, reimplemented from its paper (no code is public) and retrained on the same data, reaches 0.652 in the same detector with its published in-phase input and 0.849 when given the same I/Q input as our encoder.