Computer Vision
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Thursday 3 September
The joint interpretation of metabolic function and anatomical structure is essential for clinical diagnosis in whole-body PET/CT. Although recent advances in 3D medical vision-language models have demonstrated remarkable progress, current efforts are limited to regional CT imaging, leaving a critical void in comprehensive whole-body PET/CT analysis. In this work, we introduce MetaStructAtlas, a large-scale dataset for grounded whole-body PET/CT interpretation that synthesizes multimodal imaging with integrated anatomical, metabolic, and semantic annotations. MetaStructAtlas provides 490 co-registered 3D PET and CT volumes with 50,470 organ-level segmentation masks and grounded radiology reports. To facilitate interactive reasoning, we further developed MetaStructVQA, a standardized 3D grounded visual question-answering benchmark containing 100,565 QA pairs. This framework explicitly links diagnostic queries to visual evidence across modalities, encompassing anatomical, morphological, and metabolic characteristics. Finally, we evaluate state-of-the-art 3D medical VLMs on MetaStructVQA, establishing a robust foundation for multimodal representation learning and integrated whole-body reasoning in nuclear medicine.
Computer Vision
Semantic-Aware Subgraph State Space Model for WSI Classification in Histopathology
Histopathological subtyping relies on the recognition of characteristic histological patterns. These patterns may be expressed by individual tissue structures or by the spatial distribution and co-occurrence of multiple structures, and they often span irregularly shaped tissue regions, termed semantic units in this work. However, conventional patch-based representations may fragment such units and fail to explicitly preserve their internal spatial organization, while efficiently modeling relationships among numerous spatially separated units remains challenging. To address these limitations, we propose the Semantic-Aware Subgraph State Space Model (SASG-SSM), a flexible and efficient framework for whole slide image (WSI) classification. Semantic-Aware Subgraphs (SASGs) first approximate irregularly shaped semantic units by adaptively grouping spatially connected patches guided by class-agnostic visual-semantic priors. By representing patches as graph nodes with adjacency edges, SASGs preserve their internal spatial organization rather than treating them as an unordered set. A Subgraph State Space Module (SG-SSM) subsequently combines a graph neural network encoder for intra-subgraph topology encoding with a Mamba-based state space encoder for efficient contextualization across large numbers of subgraphs. This module integrates local structural information within semantic units with global contextual information arising from their distribution and co-occurrence across the WSI, while efficiently modeling a large number of spatially distributed regions. Extensive experiments across four WSI subtyping datasets demonstrate consistent advantages over representative state-of-the-art methods. Further evaluations under small-cohort and few-shot settings demonstrate robustness and data efficiency under limited training data. Code will be released at https://github.com/HLSvois/SASG-SSM.
Read the full abstract
Computer Vision
ToPO: Token-Oriented Preference Optimization for Text-to-Image Diffusion Models
Diffusion models have achieved remarkable success in text-to-image generation, yet aligning their outputs with human preferences remains challenging. Diffusion-DPO learns from image preference pairs, but their global labels do not specify how supervision should be allocated across spatial locations and denoising timesteps. The semantic roles of prompt tokens are also not explicitly considered when assigning these local weights. To address these limitations, we propose Token-Oriented Preference Optimization (ToPO), which builds a token-conditioned spatial-temporal weighting map for each pair. A frozen reference denoiser obtains spatial and temporal weights from local noise-prediction error differences under shared noise, while preferred-branch cross-attention links spatial evidence to content tokens. The map reweights Diffusion-DPO and supports an auxiliary pixel-midpoint ordering loss. It requires neither local preference annotations nor a separately trained reward model. Extensive experiments across multiple preference and compositional evaluations show that ToPO achieves state-of-the-art performance among the compared methods with only 500 training updates, demonstrating the potential of token-conditioned preference allocation for high-quality diffusion alignment within a limited update budget.
Read the full abstract
Computation and Language
A Circuit for Plural Reference: How LLMs Represent and Retrieve Singular and Plural Entities
Coreference resolution is an important task in contextual reasoning. In this paper, we investigate the mechanism for representing and retrieving singular and plural entities for plural reference. We use a combination of mechanistic interpretability and attention pattern analysis to study the process in which LLMs predict a pronoun to refer back to previously mentioned entities. Using a range of causal intervention techniques, we find a set of attention heads that are responsible for (1) representing coreference information in the input, (2) identifying entities that form a plural reference, (3) transferring the information to the component that is responsible for selecting the antecedents and predicting the pronoun. We also find that LLMs align with humans in preference for plural pronoun. Specifically, entities in a plural construction are more likely to be referred to as a plural entity if they are ontologically similar and are linked by the conjunction "and".
Read the full abstract
Machine Learning
Resolution-Aware Experimental Design under Partial Identifiability
Experimental design is commonly framed as choosing the experiment expected to provide the most information. Under partial identifiability however, persistent nuisance uncertainty can make the same observation carry different structural meanings. We introduce Resolution-Aware Experimental Design (RAED), which selects an experiment by the smallest expected nonempty structural candidate set achievable subject to false-exclusion control. We prove an exact cross-nuisance aliasing separation: an experiment can be preferred by structural and full-latent information gain, average classification, and nuisance-marginalized informativeness while having arbitrarily poorer valid structural resolution. RAED nevertheless preserves the expected ordering under a genuine composite Blackwell comparison. To make this criterion operational, we develop a learned score-based implementation with finite-sample nuisance-average and positive-tail calibration, and characterize a rare-tail sample-complexity obstruction. Under constrained sensing, two subsurface-flow benchmarks exhibit genuine RAED–expected-information-gain (EIG) experiment-selection disagreements, with the clearest and largest held-out resolution differences in WCA. In a fluvial benchmark, tail protection changes the selected physical experiment and replaces hard-region false exclusions primarily with explicit ambiguity. In a mechanistic methane-oxidation benchmark, a prospectively specified 5% false-exclusion tolerance also yields a nontrivial finite-sample population guarantee for tail-sensitive nuisance risk, with 95% joint confidence across all three structural families.
Read the full abstract
Data Structures and Algorithms
A PTAS for Non-Adaptive Stochastic Top-$k$ Sum under General Combinatorial Constraints
We study non-adaptive selection of a feasible set $S$ that maximizes the expected sum of the $k$ largest realized values among independent nonnegative discrete random variables. The same objective arises in team hiring and as VCG welfare in an $\ell$-unit auction. The main setting is a fixed-dimensional nonnegative packing family, whose natural LP has $d=O(1)$ packing inequalities with binary coefficients. We give a PTAS for every $k\ge 1$ on every such family, including binary one- and two-dimensional knapsack, by approximating the occupancy functional $p\mapsto\mathbb{E}[\min(k,N(p))]$ and realizing the resulting signatures in the packing LP. As a generic guarantee the scheme is essentially optimal: there is no FPTAS that works for every such $\mathcal{F}$ unless $P=NP$, and no EPTAS unless $W[1]=FPT$. An incomparable sufficient condition is a query-weight exact-sum oracle (DAG paths, matchings), which likewise yields a PTAS for every $k$. The same signatures give a PTAS for $\min_{S\in\mathcal{F}}\mathbb{E}[\mathrm{Top}_k(S)]$ on every fixed-$d$ covering family; two-dimensional covering knapsack rules out a generic FPTAS on that class.
Read the full abstract
Strongly Correlated Electrons
Hidden Magnetic Octupolar Order driven by Spin-Orbit coupling in KFeF$_3$
Phase transitions are typically driven by symmetry-breaking structural distortions that lift electronic degeneracies, yet in some cases, these transitions may be driven by a hidden order without conventional structural signatures. Here, we demonstrate that the 3$d$ perovskite, KFeF$_3$, hosts such a hidden instability. Using a combination of high-resolution powder diffraction, magnetometry, symmetry-based analysis and first principles calculations, we reveal that while the 90 K cubic to rhombohedral transition arises from conventional magnetostriction accompanying antiferromagnetic order, a second transition at 40 K lowers the symmetry to monoclinic without any detectable Jahn-Teller distortion or translational symmetry-breaking. Symmetry-based Landau analysis supports a hidden zone centered magnetic octupole order parameter, whose improper coupling accounts for the weak ferromagnetism and pronounced symmetry-breaking strain. Density functional calculations show that spin-orbit coupling suppresses the competing Jahn-Teller instability, and thereby favors higher rank magnetic multipolar degrees of freedom. These findings establish KFeF$_3$ as a model system in which the competing energy scales between spin-orbit coupling and orbital degeneracy result in the emergence of hidden, multipolar order. We show that the control of such magnetic, multipolar order could, in principle, provide a route to induce altermagnetism.
Read the full abstract
General Relativity and Quantum Cosmology
Long-lived ringing of Bronnikov–Kim brane-world black holes and wormholes
Massive-field perturbations on black-hole and wormhole backgrounds can form long-lived quasinormal ringing; this behaviour is not confined to brane-world spacetimes, but it is also not universal across compact-object geometries. Recent work on a brane-world black hole reported an unusual limiting behaviour in which the real oscillation frequency tends to zero as the scalar mass grows. We test how general this phenomenon is in the Bronnikov–Kim effective brane-world geometries, treating both the black-hole and wormhole sides of the same families. The comparison uses two complementary numerical treatments: Leaver's continued-fraction method for the black-hole side, after reducing the wave equation for the second Bronnikov–Kim family to rational form, and Chebyshev pseudospectral collocation for regular wormhole throats. The black hole of the second Bronnikov–Kim family shows the standard quasi-resonant pattern: the damping rates of the scalar branches considered here become very small, but the real frequencies remain finite. Representative wormholes from the first and second Bronnikov–Kim families also develop increasingly long-lived massive scalar modes, and extending the second-family wormhole scan to larger field mass reduces the damping by more than a factor of four relative to the massless case. Thus the robust brane-world effect found here is the appearance of long-lived massive-field ringing in both black holes and wormholes, whereas the vanishing of the real frequency is metric- or branch-dependent.
Read the full abstract
Discrete Mathematics
Nearly Tight Bounds for Proportional Group Fair Divisions and One-Sided Discrepancy
This paper studies the problem of fair division of indivisible goods among $k$ groups of $n_1,\ldots, n_k$ agents. We look at the worst downward deviation $\textit{PROP}(n_1,\ldots, n_k)$ of an agent in a group from its $1/k$-share. We improve the bounds of (Manurangsi and Meka, 2026) and show that $\textit{PROP}(n_1,\ldots, n_k) = \tilde{\Theta}(\sqrt{n/k})$, where $n = n_1 + \ldots + n_k$ is the total number of agents. For the proof of the upper bound, we develop novel discrepancy-type tools and, in particular, a way to efficiently work with one-sided discrepancy constraints.
Read the full abstract
Robotics
WISE: World-model-guided Imagination Scheduling for Efficient Post-training of Vision-Language-Action Models
Post-training VLA policies typically rely on supervised fine-tuning with costly expert demonstrations or reinforcement learning with expensive and potentially unstable real-world exploration. World models offer a promising alternative by evaluating candidate behaviors through imagined futures, yet effective post-training requires more than accurate prediction: imagination must be scheduled where it is useful, bounded within reliable horizons, and translated into trustworthy policy supervision. In robotic manipulation, the value of imagination varies substantially across execution stages, while extended rollouts can accumulate prediction errors and introduce unreliable learning signals. We introduce WISE (World-model-guided Imagination Scheduling for Efficient Post-training of Vision-Language-Action Models), a unified framework that coordinates when and how world-model imagination is used during policy refinement. WISE selectively invokes imagination at interaction-relevant states, performs bounded multi-view rollouts, evaluates candidate futures using progress and completion signals, and uses their relative outcomes to refine actions generated from real interaction contexts. Extensive experiments with both $\pi_0$ and $\pi_{0.5}$ demonstrate consistent improvements across diverse manipulation tasks while reducing GPU computation time by approximately 80% compared with full imagination. Real-world evaluations further show substantial gains in robustness and generalization under diverse real-world distribution shifts.
Read the full abstract
Computer Vision
DropClick: Semi-Automated One-Click Segmentation for Agricultural Robotic Data
Labelling vision datasets, especially for segmentation tasks, is a laborious and costly process that stymies novel developments in agricultural robotics. In this paper, we present DropClick, a click-guided segmentation tool that simplifies the annotation process. Our system utilises single-click inputs on objects to generate pseudo-labels, which can replace manual annotations. DropClick stands out as it is a semi-automated approach and does not require a click for every object in the scene. It can therefore further reduce the required amount of user input drastically. We evaluate our method on two challenging agricultural robotic datasets, SB20 and BUP20 for plant and fruit segmentation, respectively. DropClick is first trained on a small subset of just 5 images from the original training data. This DropClick model can then be deployed as a one-click segmentation system and achieves comparable or higher performance than other one-click methods achieving an mIoU of 70.0 and 72.6 points, for SB20 and BUP20 respectively. DropClick then excels at maintaining high performance when clicks are not given (e.g. dropped); when 50% of the clicks are missing it still maintains an mIoU of 68.9 and 71.3 points, for SB20 and BUP20 respectively. We validate DropClick as a pseudo-labelling approach by taking its outputs to train a Mask2Former instance-based segmentation model in a semi-supervised manner. In this process, partially removing user input from DropClick yields similar high performance when compared to providing all clicks, at 70.1 vs 70.7 points AP50 for SB20 and no difference for BUP20 at 77.0 for both models; at the same time saving 46.3% of total input for SB20 and 31.9% for BUP20.
Read the full abstract
Mesoscale and Nanoscale Physics
Weak-coupling theory of half-metals and other fractional metallic phases in biased doped Bernal bilayer graphene
The paper presents a theoretical study of many-body electronic phases in doped and electrically biased Bernal-stacked (AB) bilayer graphene. We develop a variational mean-field theory with no fitted parameters. For the electron-electron interaction, we employ a parameter-free random-phase approximation to model the short-range screened Coulomb repulsion. The remaining long-range Coulomb potential energy is dictated by the geometry of the sample, behaving as a parallel-plate capacitor. We formulate the theory directly in terms of the experimentally controlled displacement field $\frak D$ rather than the interaction-renormalized interlayer potential difference. It makes our theory better suited for direct comparison with experiment. The resulting phase diagram is quite rich. It hosts several fractional-metal states connected by first- and second-order transitions. At higher bias and doping, three distinct fractional metallic states emerge. We classify these phases by the number of doped sectors and the symmetry of the order parameters. Our results - obtained without any parameter fitting - reproduce key qualitative and quantitative features of recent experiments. This includes the energy scale associated with the loss of fractional metallic order. At lower bias and doping, the model stabilizes a broad spectrum of fractional-metal phases with more exotic symmetry-breaking patterns. This suggests that further experimental exploration of the latter regime is warranted.
Read the full abstract
Human-Computer Interaction
Exploratory Unstructured Data Analysis: A Formative Study and Implications for Human-AI Collaboration
We propose a conceptual framework for exploratory data analysis of (large) unstructured data (EluDA), combining classical elements (querying, visualization) with active knowledge construction in the "search for structure". In a formative study, users conceptualized a structure for an image dataset during exploration. We found that users conceptualize by building faceted classifications bottom-up and rarely create meaningful spatial categorization during this process. We also evaluated CLIP for zero-shot assignment and semantic categorization, finding that it remains unreliable for assigning user-defined concepts to images but does support semantic grouping. Based on these findings, we identify and discuss four key opportunities for human-AI collaboration in EluDA: intelligent sampling and visualization to maximize data visibility; incremental and few-shot learning to minimize effort for reliable assignment; automatic category, concept, and facet suggestions to reduce effort during the search for structure; and the necessity for effective trust calibration methods.
Read the full abstract
Computer Vision
Understanding Autonomous Driving Datasets by Describing Differences between Image Subsets in Natural Language
Understanding the composition of large-scale autonomous driving datasets is essential for safety, robustness, and reliable operation across domains. For example, domain shift between locations could lead to the operating environment being misaligned with the training data, resulting in potentially dangerous performance degradation. Yet, existing data analysis pipelines largely rely on metadata, predefined labels, or manual inspection, which provide limited semantic insight or do not scale. This paper studies set difference captioning: given two subsets of images, the goal is to produce a natural-language hypothesis describing differences between the target and reference set. Building on a two-stage formulation, we adapt the method to autonomous driving by focusing on object-centric patches derived from object detection, which simplifies aggregation and enables attribution of differences to specific object instances or categories. To evaluate this setting in-domain, we introduce a new benchmark, AD-Diff Bench. Low-concentration experiments assess the suitability of set-difference-captioning approaches to sparse, real-world differences. We restrict our experiments to open-weight models to support reproducibility and ease of deployment. The proposed benchmark and analysis provide a step towards practical, human-interpretable dataset introspection for autonomous driving datasets. Our implementation and benchmark dataset are available at https://github.com/KIT-MRT/AD-Diff
Read the full abstract
Differential Geometry
Quadratic Scalar Curvature Decay and Uniform Positivity in Dimensions Four through Seven
Let $4\le n\le7$ and let $(M^n,g)$ be a complete, connected, orientable, noncompact Riemannian manifold of positive scalar curvature. We prove that if the asymptotic quadratic scalar curvature coefficient of $g$ is greater than $(n-1)/n$, then $M$ carries a complete smooth metric whose scalar curvature is at least one. The threshold $(n-1)/n$ and the strict inequality are optimal. This confirms the second part of Gromov's critical rate of decay conjecture in dimensions four through seven.
Read the full abstract
Computer Vision
CoFiE: Coarse-to-Fine Evidence Selection for Efficient Streaming Video Understanding
Streaming video understanding requires Vision Language Models (VLLMs) to process growing video streams and answer user questions under tight latency constraints. Existing methods improve efficiency through token pruning and memory-bank schemes, but mainly reduce visual tokens after visual encoding. Consequently, downstream token pruning alone cannot substantially reduce end-to-end latency because the expensive frame encoding cost has already been incurred. We propose CoFiE, a Coarse-to-Fine Evidence Selection framework that decouples evidence selection into a coarse, query-agnostic filtering stage before the vision encoder and a fine, query-specific refinement stage during LLM prefill. CoFiE introduces Novelty-Guided Frame Filtering to retain visually distinctive candidate frames and Query-Specific Evidence Refinement to select the frames most relevant to the user query. This design removes substantial redundancy before frame encoding while preserving query-specific refinement once semantic information becomes available. Experiments show that CoFiE establishes a new state-of-the-art accuracy-efficiency trade-off across multiple video understanding benchmarks, reaching 78.86% accuracy on StreamingBench and 68.72% on OvO-Bench, with improvements of up to 3.15% over prior methods. Even with up to 80% evidence-frame filtering, CoFiE outperforms strong open-source multimodal models while improving end-to-end inference latency by up to 2.54 times.
Read the full abstract
Information Retrieval
LLM4AIGQ: LLM-based AI Guidance Query Generation Framework for Multi Interest Mining
Guidance queries stimulate user consumption by extracting preferences to provide search queries with guidance value, playing a crucial role in the e-commerce field. Traditional AI-generated queries (AIGQ) generation primarily relies on a two-stage "Query-to-AI-Generated-Query" (Q2AIGQ) association paradigm, first recalling user primary search queries from user profiles, historical behavior sequences, item-side information, and the current query through multi-path retrieval, then generalizing AIGQ via rule-based methods. This approach suffers from semantic drift due to information cascade loss; additionally, primary search query derivation heavily depends on "user-item" co-occurrence relationships, lacking exploration of user multi-interests, resulting in guidance queries with low value and mismatched purchase intent. To address the expressive limitations of traditional co-occurrence-based retrieval, we propose LLM4AIGQ, an LLM-based solution for generating AI guidance queries tailored to users' multi-interests. This approach segments user interests by integrating user profiles and historical interaction sequences, infers specific consumption intents for each sub-interest, and subsequently generates corresponding AIGQ. In terms of model training, we employ a post-training pipeline comprising Supervised Fine-Tuning (SFT), Reinforcement Learning (RL), and Direct Preference Optimization (DPO) to enhance the model's capability in generating AIGQ. We also introduce a multi-level reward design to satisfy the requirements of multi-objective optimization and long-chain reasoning in practical applications. Regarding deployment, we adopt a nearline-generation and online-read architecture to meet latency constraints. Extensive experimental analyses demonstrate that our model achieves robust performance in both offline evaluations and online A/B tests.
Read the full abstract
Computer Vision
Do Video Generators Track the World Across Segments? A Benchmark and Method for World-State Reasoning in Video Continuation
Video generators build long videos by composing shorter parts, either by generating segments one after another or by autoregressively extending chunks. Each new part usually depends on memories of historical observations, such as recent frames, selected key frames, memory banks, or cached features. These memories preserve visible evidence from the past, but current generators do not reliably turn such evidence into a world-state interface: what holds in the video world after previous actions and how it should change under the next prompt. A past frame remains valid history, but it may not describe the state needed by the next segment; some states must instead be inferred from occluded or implicit changes rather than copied from a directly observed frame. This creates a simple but overlooked question for video continuation: given a previous video, its prompt, and a new prompt, can a model generate a continuation that reflects the state determined by both the historical video and the new prompt? To answer this question, we introduce Statebench, a benchmark that targets this gap by testing continuations over three state categories: past-visible states, occluded-process states, and complex-transition states. We further propose Stateagent, which explicitly maintains an entity-state representation, updates it under the new prompt, grounds the predicted post-action state as a future end frame, and renders the next video. Experiments show that our method improves controlled video continuation by raising the all-case state score (SCS-All) from 45.2 to 69.3, and also benefits story generation at the one-minute scale. Code is avaliable at https://github.com/AMAP-ML/StateAgent.
Read the full abstract
Databases
Convolution Sum-Product Queries
We study query evaluation for an extension of sum-product queries (SPQ) that allows atoms with linear combinations of variables (e.g. $A(2X+3Y-Z, Y+Z)$), which we call convolution sum-product queries (CSPQs). These queries arise in both practical settings (e.g. image processing workloads) and theoretical ones (e.g. the $(\min,+)$-convolution and $k$-SUM conjectures), and capture SPQs with added linear equality constraints. While prior work has considered the impact of linear in- and dis-equality constraints on query evaluation, the techniques developed in that setting are asymptotically sub-optimal for CSPQs. To address this, we describe several evaluation algorithms for CSPQs that leverage linear algebra techniques like rank analysis, quotient spaces, and variable substitution. First, we adapt Worst-Case Optimal Joins to CSPQs, achieving a runtime similar in spirit to that for conjunctive queries. Then, we extend the definition of tree decompositions (TDs) to CSPQs, and describe a factorized execution. In this extension, linear combinations are first-class citizens and play the same role as variables in traditional TDs. We define three width measures that bound the complexity of this execution with respect to the size of the domain values, of the active domain, and of the relation's support. Lastly, we show how these methods can be applied to arbitrary fields beyond the rationals.
Read the full abstract
Optimization and Control
Bi-objective Two Phase Transportation Problem
This paper discusses a bi-objective two-phase transportation problem (BTPTP) in which transportation takes place in two phases. The whole set of source-destination links is partitioned into two disjoint sets namely, Phase-I and Phase-II. The transportation among the source-destination links of Phase-II set is done only after the transportation among the source-destination links of Phase-I set is complete. The transportation among the source-destination links in each phase is done in parallel. The problem discussed in this paper concentrates on minimizing two objectives viz., sum of transportation times and sum of transportation costs of both the phases, simultaneously. Due to conflicting nature of the objectives, a polynomial time iterative algorithm named as BTPTP-Algorithm is developed that finds all of its non-dominated solutions. At each iteration, the BTPTP-Algorithm solves some specific restricted cost minimizing transportation problems and finds corresponding optimal feasible solutions. The restrictions imposed on the source-destination links of Phase-I and Phase-II sets, in each of these problems, depend upon their corresponding Phase-I and Phase-II transportation times obtained at the previous step. The algorithm is developed in such a way that it expels all the dominated solutions and records all the non-dominated solutions of the problem, systematically. A proof of the BTPTP-Algorithm's capability of recording all the non-dominated points is provided. Further, a numerical illustration is given in support of the theory and computational behavior of the BTPTP-Algorithm for various BTPTP instances of different sizes, is provided in terms of CPU time.
Read the full abstract
Logic in Computer Science
The Stochastic Target Discounted-Sum Problem
The target discounted-sum problem (TDS) asks, given a finite integer alphabet $\Sigma$, a rational discount factor $\lambda$, and a rational target $t$, whether some infinite sequence over $\Sigma$ has discounted sum exactly $t$. This problem remains open and underlies several open questions in automata theory, games, and Markov decision processes. We introduce and solve its stochastic counterpart, the stochastic target discounted-sum problem, which replaces existence by computation of the probability. We show that the probability that a random sequence generated by a finite Markov chain has discounted sum $t$ is rational and computable in pseudo-polynomial time. We further show how to decide, in polynomial time, whether the discounted-sum distribution of a Markov chain is atomless, and how to approximate to an arbitrary precision the probability that the discounted sum exceeds a rational threshold. Our techniques for the stochastic TDS problem allow us to make progress on TDS objectives in stochastic games, which are known to be as hard as the TDS problem. Restricting the maximizing player to finite-memory strategies, while allowing the minimizing player to use arbitrary strategies, we reduce the value problem and the synthesis problem to corresponding problems for safety objectives in stochastic games. This yields computable optimal values and deterministic optimal strategies with pseudo-polynomially bounded memory for stochastic games, and results in pseudo-polynomial-time algorithms for special cases of Markov decision processes and deterministic two-player games.
Read the full abstract
Probability
Volterra Integral Reduction for Boundary Diffusion Problems
This paper addresses a class of integral representations of the form \begin{equation} f(t,x)=g(t,x)+\int_0^t k(t,s)\, p(t-s,x,y)\, \partial_x f(s,y^+)\,ds, \qquad 0 \le t \le T, \end{equation} where $f$ is unknown, $p$ is the transition density of a diffusion process, and $g, k$ are prescribed functions. For an arbitrary diffusion process with sufficiently regular coefficients, we prove that this problem is equivalent to a Volterra integral equation of the second kind. This reduction provides a unified framework for both theoretical analysis and numerical approximation. An example of the implementation in the context of financial mathematics is presented.
Read the full abstract
Computer Vision
ARCOS: Zero-shot Boundary Localization for Corneal Layer Segmentation Across Optical Coherence Tomography Devices
Accurate segmentation of corneal layers in optical coherence tomography (OCT) is essential for quantitative assessment of corneal morphology, including layer thickness and structural changes associated with disease or surgery. However, automatic segmentation remains challenging because corneal interfaces are thin, affected by speckle noise, and variable across acquisition devices. In this work, we propose ARCOS, a patch-based zero-shot boundary localization framework for corneal layer segmentation in clinical anterior-segment OCT images. Rather than performing conventional region classification, the method predicts boundary heatmaps for the main corneal interfaces from overlapping native-resolution patches. Patch-level predictions are stitched across the full B-scan and converted into boundary locations to obtain continuous, anatomically ordered layer segmentations. The network combines multi-scale feature fusion with a self-conditioned refinement module that uses intermediate boundary information to improve local heatmap predictions while preserving spatial detail. The method was evaluated on clinical OCT images acquired from multiple devices and compared with representative segmentation baselines using boundary localization and derived thickness metrics. The proposed method achieved an off-by-one boundary localization accuracy of 95.1% and a mean absolute boundary error of 0.514 pixels on the matched-device test set. In zero-shot cross-device evaluation, it maintained an average off-by-one accuracy of 84.3% and a mean absolute boundary error of 0.855 pixels across unseen acquisition devices, outperforming the baseline models. Thickness estimates derived from the predicted boundaries showed low error across corneal regions, supporting the method's use for quantitative corneal OCT analysis.
Read the full abstract
Machine Learning
Out-of-Distribution Generalisation with Sequence Models in Offline Multi-Agent Reinforcement Learning
Generalising to unseen tasks remains a fundamental challenge in offline multi-agent reinforcement learning (MARL). In this work, we present a principled analysis of zero-shot task generalisation in the offline setting and conduct an extensive empirical investigation into the scaling behaviour governing task diversity, dataset size, and network capacity. To facilitate this study, we extend offline sequence modelling architectures to handle multi-task observation and action spaces alongside variable agent counts across tasks. Our primary finding is that scaling task diversity—rather than sheer dataset size is the dominant factor in achieving robust zero-shot transfer. Through large-scale experiments across four challenging environments (Connector, RWARE, SMAX, and LBF), we demonstrate that our multi-task approach achieves a mean improvement of 3.2x on held-out test tasks compared to single-task models and consistently outperforms strong behaviour cloning baselines. These results suggest that the development of generalisable MARL agents should prioritise the diversity of the training distribution with varying numbers of agents, providing a roadmap for scaling offline MARL effectively.
Read the full abstract
Computers and Society
WebXR and Commercial Game Engines for the Metaverse: A Socio-Technical Analysis of Openness, Interoperability, and Sustainability
The Metaverse is often framed as a persistent, interoperable, and embodied network of virtual and augmented environments. Yet, most contemporary XR applications are developed through commercial game engines and distributed through proprietary app stores, creating tensions between openness and platform dependency. This paper critically examines open WebXR technologies with conventional commercial game-engine pipelines, with particular attention to XR hardware, software architectures, developer workflows, governance, ethics, interoperability, and sustainability. We argue that WebXR may provide a viable route toward a more accessible, device-independent, and institutionally sustainable Metaverse, especially for education, research, cultural heritage, prototyping, and public-interest applications. At the same time, commercial engines remain advantageous for graphically intensive, low-latency, deeply integrated, and large-scale XR products. The paper concludes that the choice should not be framed as WebXR versus engines, but as a continuum: WebXR is preferable when accessibility, interoperability, low-friction deployment, and long-term maintainability are primary goals, whereas native engines remain preferable when performance, platform-specific hardware access, and production-grade tooling dominate.
Read the full abstract
Human-Computer Interaction
PlanePivoting: Exploration and Optimization of Gaze-Mouse Cursor Alignment for Spatial Object Translation
As XR matures into a ubiquitous computing platform, the disconnect between 2D and 3D input modalities remains a critical barrier to seamless workflow. Frequent transitions between the mouse for 2D precision and hand gestures for 3D manipulation induce significant physical fatigue and cognitive load. To address this, we introduce PlanePivoting, a multimodal interaction technique that extends standard mouse input into 3D space by leveraging gaze-mouse alignment. This technique dynamically modulates the translation plane based on the spatial overlap between the gaze and mouse cursor, eliminating the need for physical input modality switching. To systematically explore the foundational design space of gaze-mouse coordination and optimize key variables, we conducted a user study comparing PlanePivoting with a standard 3D Gizmo interface across two translation mapping profiles and two gaze cursor apertures. Results demonstrate that PlanePivoting outperforms the Gizmo on efficiency metrics while maintaining comparable precision and yielding higher subjective satisfaction. This study demonstrates the potential of gaze-mouse alignment for efficient spatial manipulation between 2D and 3D environments.
Read the full abstract
Atmospheric and Oceanic Physics
Cold Extremes during Dansgaard-Oeschger Oscillations
This paper studies the statistics of extreme cold air surface temperatures in a climate that experiences abrupt changes. We employ a CCSM4 simulation of Last Glacial Maximum conditions that exhibits rapid switching between stadial and interstadial states as data. Non-stationary linear Generalized Extreme Value (GEV) distributions are fitted to find the regions that show the most prominent changes and associate them with physical processes. The results are then compared with a non-linear model, which gives a more detailed picture of how the parameters change as a function of the AMOC strength. While different regions of the world show different extremal behaviours, many regions show an approximately linear relationship between the parameters of the GEV distributions and the strength of the AMOC within the stadial and interstadial states, with some non-linear oscillation or jump where the transition between them takes place. Physical processes such as the expansion and retreat of the sea ice and the relative changes in the strength of the currents are shown to impact the magnitude and variability of the extremes, with significant changes observed in the three parameters of the GEV. They also create teleconnections that are compared whenever possible with various proxies for the temperature of the air and the water. We show how mapping the parameters of the GEV distributions into the AMOC strength gives a way to compare results between different climate models and different climate states. Comparisons, however, need to pay heed to the dynamical characteristics of the state and the location to be meaningful.
Read the full abstract
Computer Vision
Cross-Dataset Transfer and Reliability of Explainable Artificial Intelligence for RhythmFormer Remote Photoplethysmography
Background. Remote photoplethysmography estimates the cardiovascular pulse from facial video, and its explanations have rested on inspecting heatmaps rather than on quantitative evidence about where a model reads it. We quantified the explanations and asked whether such explanations transfer between datasets and track model performance. Method. We trained eight condition-specific RhythmFormer models on NCKU-rPPG, recorded under three illumination levels, speaking, rotation, and cycling, estimated one heart rate per 5.12-second clip, and set them beside a UBFC-rPPG reproduction. Raw attention, rollout, attention flow, and Beyond Intuition were assessed by skin coverage and the Salience-guided Faithfulness Coefficient (SaCo). Results. Beyond Intuition ranked highest on both datasets, at median coverage 0.789 and SaCo 0.837 on Static level 3 against 0.826 and 0.917 on UBFC-rPPG; lower ranks differed. Within one participant of one condition, neither measure was related to a clip's heart-rate error, waveform correlation, or signal-to-noise ratio on either dataset: 186 of the 252 coefficients fell below $|\rho|=0.10$ and 28 reached $p<0.05$ against the 13 expected by chance. Across the eight scenarios only Beyond Intuition's coverage followed the three performance measures, at $\rho=-0.43$, $+0.57$, and $+0.43$, while the attention-only methods' SaCo ran opposite to each. It failed at 40 lux alone, its median coverage falling to 0.180 and its median SaCo to $-0.178$, whereas motion degraded the estimates far more without such a drop. Conclusions. Skin coverage and SaCo carry information complementary to the performance measures rather than a proxy for them: attributing to the skin does not guarantee an accurate estimate. What an attribution reveals about a condition is where the model looks rather than how faithfully its map is ordered.
Read the full abstract
Machine Learning
Extracting Forgotten Prompts from Targeted Unlearned Models
Recent unlearning methods (e.g. NPO, DPO, LUNAR) make use of refusal alignment to suppress forgotten data. However, it has been shown that refusal responses might leave traces of unlearning, and recent attacks have been able to successfully recover some of the unlearned knowledge. In this paper, we uncover a new vulnerability. Existing attacks typically assume that the forgotten prompts are already known to the adversary and focus on recovering their answers. However, we show that the forgotten prompts themselves can be extracted by using the retained data and black-box access to the model. Our attack, Targeted Active Search (TAS), first identifies the forgotten entities by constructing canonical templates and entity pool, and selectively querying the model using the most informative template-entity pair under a limited query budget. Once the entities are identified, TAS instantiates prompt templates with those entities to probe the unlearned model and reconstruct the forgotten prompts. Experiments across three unlearning methods with three datasets and three LLMs shows that TAS recovers the forgotten entity with $100\%$ accuracy and reconstructs up to $95\%$ of forgotten prompts, all while using up to $99.7\%$ fewer queries than naive probing.
Read the full abstract
Human-Computer Interaction
Point&Spawn: Mid-Air Reference-Free Object Instantiation Using Gaze and Hand Gestures in Extended Reality
Mid-air object instantiation in XR requires users to specify a 3D position without spatial references, such as surfaces or existing objects. We present Point&Spawn, a staged pipeline for pre-instantiation position specification through Direction Setting, Depth Setting, and Position Refinement within a continuous gesture flow. We evaluated six controller-free techniques combining Gaze or Non-Dominant Hand (NDH) direction setting with Ray Intersection, Relative Gain, or Drag&Hold depth setting in a user study (N=24) across Near and Far spawn depths. Relative Gain and Drag&Hold yielded faster and more accurate spawning, lower workload, higher usability, and greater preference than Ray Intersection. The shoulder-referenced NDH ray improved speed and coarse accuracy, whereas the viewpoint-based Gaze ray reduced hand movement with comparable final accuracy. Farther spawn depth imposed greater temporal costs as well as Gaze and accuracy costs with Ray Intersection. These findings offer empirical guidance for designing direction and depth control in spawning in XR.