Signal Processing
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Saturday 7 November
High-fidelity radio map prediction is essential for wireless network optimization, yet existing single-network models frequently fail to generalize across diverse multi-band and multi-terrain scenarios due to rigid parameter constraints. We introduce LLM-MoE, a framework that employs a divide-and-conquer Mixture-of-Experts strategy to dynamically allocate modeling tasks to matched expert sub-networks. Our approach features a heterogeneous expert pool for characterizing varied physical propagation behaviors, combined with an LLM-driven semantic extractor that provides global physical awareness. A cross-modal fusion module integrates these macroscopic semantics with local spatial features for adaptive routing. Experimental results on the SpectrumNet dataset show that LLM-MoE outperforms top-tier baselines, achieving a 10.0% reduction in Mean Squared Error (MSE) and 9.7% in Normalized Mean Squared Error (NMSE), while ensuring robust generalization across heterogeneous environments.
High Energy Astrophysical Phenomena
SMMAN studies of AGN variability. II. Radio and optical quasi-periodic oscillations and changing variability states in the relativistic jet of the blazar PKS 0754+100
We report the detection of $\sim$280-day quasi-periodic oscillations (QPOs) in the SMMAN observations of PKS 0754+100 at 4.8 and 23.6 GHz, which may also be present in VLBI observations made at 15 GHz. The onset of a transition of the amplitude of variability in the radio flux density, observed more clearly in the 15 GHz and 23.6 GHz observations, coincides with the detection of a likely QPO of $\sim$1.4 days in optical observations made by the Transiting Exoplanet Survey Satellite. This transition in radio variability also occurs close in time to the disappearance of the secondary knot in the VLBI radio map during the lower variability state. An X-ray flare was observed 100 days before this transition, suggesting that the mechanisms causing these quasi-periodicities are related. After the transition to a higher variability state, both the 15 and 23.6 GHz light curves exhibit indications of QPOs at $\sim$400 days. For the first time, a QPO detected in the radio band from a BL Lac has been seen to appear to shift to a QPO of lower frequency, with that transition occurring near the appearance of a QPO in optical emission at a much shorter period. This multi-waveband behavior could be related to the internal helical structure of the jet, where an instability or kink develops due to magnetic reconnection and propagates outward.
Read the full abstract
Computer Vision
Understanding the Effect of Visual Attention in CNNs: A Decision-Space Alignment Perspective
Visual Attention enhances convolutional neural networks (CNNs) by modulating intermediate features, yet their effects on downstream decisions remain less understood. We present a decision-space framework that models multiplicative attention as an intermediate feature perturbation and derives a first-order prediction of its effect on the target-versus-competitor classification margin. We further introduce the Decision Alignment Score (DAS) to quantify the directional alignment between the perturbation and the local margin gradient. We evaluate the framework on CIFAR-100 using SimAM and CBAM across five popular CNN backbones, ResNet-18, ResNet-20, ResNet-50, ResNet-56, and MobileNetV2. The analysis examines prediction agreement, DAS margin correlations, and downstream activation-state changes. Across most final-attention configurations, predicted and actual margin changes show good agreement, while DAS exhibits positive correlations with actual margin changes. Single-attention experiments further show that prediction agreement and DAS margin correlations depend on attention placement, with stronger associations at later stages.
Read the full abstract
Machine Learning (Statistics)
Deep backward multistep high-order regression methods for high dimensional nonlinear parabolic partial differential equations
We propose a deep backward multistep high-order regression (MHDBR) method for semilinear parabolic partial differential equations through their backward stochastic differential equation representations. The method combines multistep time discretization with regression targets formed by conditional expectations. Under explicit regularity, starting-value, forward-approximation and network-moment assumptions, we prove an error bound separating time discretization, conditional Monte Carlo sampling and neural-network approximation. The analysis assumes exact population minimization; finite-training-sample generalization and optimization errors are not included. For generators independent of the gradient variable, suitably constrained GroupSort networks and sufficiently large conditional sample sizes give order $J$ in the stated nodal $L^2$ error norm. The required network sizes depend on dimension, so this result does not establish dimension-independent complexity. We present a reproducible time-discretization test satisfying the analytical assumptions and three broader neural-network benchmarks, whose relation to the assumptions is discussed explicitly.
Read the full abstract
Information Theory
Learning-Assisted Beamforming Optimization for Hybrid XL-RIS-Aided Communication
Hybrid active-passive extremely large-scale reconfigurable intelligent surface (XL-RIS) architectures mitigate the double-hop path loss of passive RIS while enabling near-field spatial energy focusing, but impose a computationally intensive mixed discrete-continuous beamforming design problem under strict radio-frequency (RF) power constraints. This paper investigates a hybrid XL-RIS-assisted downlink multi-user multiple-input single-output (MU-MISO) system in the near-field regime. The joint optimization of the base station (BS) precoder, discrete XL-RIS phase shifts, and active-element amplification gains is addressed via a discrete coordinate ascent (DCA)-based alternating optimization (AO) framework, in which discrete variables are updated through local index-domain searches and the BS precoder is derived in closed form via a noise-aware minimum mean-square error (MMSE) criterion. To reduce the per-sample computational cost for XL-RIS apertures, a deep unfolding (DU) framework is proposed as a low-complexity alternative, unrolling a fixed number of Riemannian gradient ascent steps on the sum-rate objective and recovering the discrete phase configuration via quantized phasor projection, with one learnable step size per layer trained in a self-supervised manner. Numerical results under near-field spherical-wave channel models confirm that the DU framework achieves sum-rate performance close to the AO scheme with substantially lower and nearly constant runtime, and that both schemes consistently outperform manifold optimization under discrete hardware constraints.
Read the full abstract
Signal Processing
Wireless Interference Recognition with Limited Data via Utilizing Diffusion Model
Accurate recognition of interference signals is essential for maintaining the reliability and security of modern wireless communication systems. However, existing approaches face several significant challenges: the high cost and the scarcity of accurately labeled datasets adaptable to diverse environments, coupled with the limited ability to evaluate uncertainty factors arising from propagation in complex electromagnetic environments. To address these challenges, this paper proposes a diffusion-driven model for wireless interference recognition. Specifically, we develop a composite data augmentation strategy that integrates geometric transformation and modal augmentation to enrich the limited data distribution. Geometric transformation expands scarce signal data, and modal augmentation generates semantically consistent multimodal data by modeling the correlations among cross-domain features. Following that, we propose a novel diffusion-based generative model with an interference classifier to accurately capture the distribution of interference signals within given signal data. It hence significantly improves the understanding of environmental uncertainty factors. Extensive experiments demonstrate that the proposed model outperforms state-of-the-art methods by over 5.84% in wireless interference recognition.
Read the full abstract
Information Theory
On Relative Greedy Weights
Relative greedy weights were introduced by Li and Liu in [19] as a measure of information leakage in coordinated two-party wire-tap channels of type II when the adversary follows a greedy strategy. Their definition can be viewed as a relative version of the CEZ greedy weights, which are only locally greedy. In this paper, we revisit the concept of greediness and relative greediness introducing new measures, and we discuss other perhaps more relevant applications of relative greedy weights than the coordinated two-party wire-tap channel of type II. We pay special attention to bottom-up greedy weights, which correspond to an adversary with an ultimate greedy approach, and we present families of nested codes for which such weights are larger than the relative CEZ greedy weights.
Read the full abstract
Robotics
Online Geometric Consistency Assessment of LiDAR Measurements via Recursive Bayesian Comparison
LiDAR-based localization and mapping systems produce registered point clouds that serve as geometric measurements for downstream estimation and navigation. Registration errors introduce duplicated surfaces and structural deformation into these measurements. In the absence of an independent reference, online assessment relies on geometry accumulated from earlier measurements. Such assessment requires tolerance to sampling variation and an accumulation process in which observations enter the history only after evaluation. We therefore formulate a recursive Bayesian comparison that couples geometric assessment with history maintenance. A Normal-Inverse-Wishart model represents historical voxel geometry and its uncertainty, while a weak isotropic reference adapts to current sampling. Comparing the marginal predictive support of the same current points under both models yields Normalized Drift Evidence (NDE). Assessment precedes assimilation: positive NDE triggers an alarm and preserves the preceding historical state; otherwise, the current observations update the history. Across 14 real-world sequences totaling approximately 73 min from four platform types, alarms identified all nine corrupted sequences and were absent from all five consistent sequences. In four representative scenarios with event references, first alarms occurred within 0.9 s of the reference times, with no preceding alarms. Cross-backend evaluation demonstrates consistent responses to reconstructed geometry. Ablation and controlled perturbation results establish the roles of uncertainty modeling, recursive history, displacement, and geometric support. Mean processing time is 4.2–6.1 ms per update at 1 Hz. In ten UAV trials, alarms initiated odometry isolation, landing, relocalization, and mission continuation.
Read the full abstract
Neurons and Cognition
Pixels over Symbols: Sensory Realism Improves Behavioral Alignment in Models of Cognition
Computational models of cognition have traditionally relied on abstract symbolic inputs that approximate sensory experiences with simplified vector representations. This simplification reflects historical computational constraints and a long-held assumption that the details of sensory realism are incidental to the study of cognitive processes. Recent advances have enabled the development of sensory-realistic models capable of performing diverse cognitive tasks, yet whether sensory realism meaningfully shapes cognition has not been formally established. We address this gap through a large-scale empirical comparison of abstract and sensory-realistic neural network models trained on a battery of memory-dependent decision-making tasks. Our results show that sensory-realistic models mirror human behavioral response patterns significantly more than their abstract counterparts, though a substantial gap to human internal behavioral consistency remains even for the best tested models. By analyzing the internal dynamics of these systems, we find that the learned recurrent weights are not the main source of this divergence. In fact, frozen abstract-trained weights driven by sensory-realistic inputs recover the human consistency of sensory-realistic models. The resulting dynamics are also within the seed-to-seed variability of sensory-realistic models. Together, our findings argue that across a broad set of visual working-memory tasks, sensory realism accounts for a substantial portion of the model-human behavioral agreement, more than the choice of sequence architecture and comparable to the choice of vision encoder. These results suggest that findings obtained from abstract task models in neuroscience should be more carefully interpreted in light of their potential behavioral limitations.
Read the full abstract
Chemical Physics
Symmetry and Redundancy in Cremer-Pople Phases for Ring Puckering
Cremer-Pople (CP) coordinates provide an exact description of ring puckering, yet direct sampling of their phases generates many symmetry-equivalent representations of the same puckering pattern. We develop a general analytical framework that identifies these equivalences before Cartesian reconstruction and geometry optimization. For homogeneous rings of arbitrary size, explicit transformations of the CP phases and the signed special coordinate of even-membered rings generate the complete symmetry orbit of the ordered elevation pattern. We classify the redundancies into four types and derive a factorized orbit-size formula linking CP multiplicity to puckering symmetry. These results yield exact single-phase fundamental intervals and sequentially reduced multiphase covering domains, including the restriction to a positive special coordinate. The framework is extended to rings containing one heteroatom or two identical heteroatoms, with reductions determined by the operations that preserve their placement. We also introduce a semidiscrete conformational nomenclature that combines symbolic amplitude distributions with explicit active phases. In an application to cyclohexadecane, symmetry reduction alone decreases the number of sampled CP points by approximately a factor of thirty. Combined with selective phase-grid coarsening for three ordinary modes, the reduction approaches two orders of magnitude. These developments provide a systematic basis for generating, classifying, and efficiently sampling ring puckering patterns in CP space.
Read the full abstract
Machine Learning
Confronting Uncertainty in Mechanism-AI Coupled Dynamical Systems via Optimal Transport and Information Geometry: Part I. Distribution Learning
Mechanism-AI coupled dynamical systems are increasingly used to study complex biological systems. However, uncertainty associated with noisy, sparse, and partially observed data limits their reliability. This paper presents a systematic mathematical formulation for treating these uncertainties from the perspective of optimal transport and information geometry. We formulate the uncertainty quantification problem as the minimization of an energy functional on a probability density space equipped with a Riemannian metric, with the loss evaluated through a differential equation, and give a unified description of the gradient flows of the energy functional under specific metrics, such as the Wasserstein and Fisher-Rao metrics. Furthermore, we present corresponding numerical methods for solving these gradient flows using abstract adjoint sensitivity analysis. The formulation connects a range of Bayesian inference methods for dynamical systems; numerical experiments compare posterior approximation and predictive uncertainty under a common computational budget. Moreover, it also provides a geometric perspective on dynamical modeling and the study of heterogeneity evolution.
Read the full abstract
Computer Vision
ZARA: Zero-shot Animatable and Relightable 3D Avatar reconstruction from a single image
Recent large human reconstruction models create remarkable animatable 3D avatars from a single image, but their appearance remains entangled with the illumination and cannot be relighted. In contrast, existing 3D relightable-avatar reconstruction methods typically require person-specific optimization. We present ZARA, a zero-shot feed-forward framework that reconstructs an animatable 3D Gaussian avatar together with its intrinsic attributes from a single image. ZARA couples a pretrained large-scale human reconstruction model with a Material Transformer to inherit geometry priors while learning material-aware Gaussian representations. To preserve localized details, we introduce Gaussian-aligned material feature fusion strategy, which injects multi-scale material features onto individual Gaussians and combines them with geometry tokens. As most human-avatar datasets lack ground-truth material maps, we use pseudo-material labels as conditioning inputs and supervision targets during training. We further introduce MatFace regularization to preserve facial structure and color in albedo prediction under unreliable facial pseudo-material supervision. Extensive experiments demonstrate that ZARA reconstructs animatable and relightable humans within tens of seconds without person-specific optimization or post-processing, outperforming existing baselines in both relighting quality and materials generalization ability. As the ability of recovering the physically-based-rendering attributes of the human, ZARA also enables direct material editing of the avatar's appearance.
Read the full abstract
Audio and Speech Processing
BAmbiDrop: Array-Agnostic Speech Enhancement with Zero-Shot binaural Reproduction using Multi-center Ambisonics Encoding
Array-agnostic enhancement networks generalize across microphone geometries but produce a single channel, while binaural enhancement methods preserve interaural cues only for one fixed device geometry. We propose BAmbiDrop, a framework that extends array-agnostic enhancement to binaural reproduction without retraining. Exploiting the fact that the Ambisonics origin is an arbitrary point in space, a frozen single-channel AmbiDrop network estimates a complex mask at multiple centers across the array; these masks are fused by a complex mean into a single shared mask, applied to the microphone signals and rendered by fixed BSM filters. Across 21 free-field and rigid-sphere arrays, the fused mask improves enhancement and interaural cue reproduction simultaneously, whereas masking each microphone independently degrades the latter without a consistent gain in the former.
Read the full abstract
Computation and Language
Large Language Models without QK-Normalization
Query-Key Normalization (QK-Norm) is widely adopted in modern large language models to control the growth of attention logits and improve training stability. However, query and key norms are not merely sources of instability. The query norm controls the sharpness of the attention distribution, while the key norm determines how strongly a token attracts different queries. These norms therefore carry token-specific information about attention selectivity and token importance. By normalizing both, QK-Norm suppresses this information. This raises a natural question: can we control query-key representations without discarding their sensitivity to magnitude? To address this question, we propose QK-MAP, a Magnitude-Aware Projection for constructing queries and keys. Rather than removing magnitude through normalization, QK-MAP transforms the input representation using multiple learnable frequencies, so that changes in magnitude are encoded as bounded phase patterns. This allows magnitude to be represented through phase and direction rather than vector length alone. Its periodic features have a fixed norm while remaining sensitive to magnitude changes, separating magnitude encoding from norm growth. Multiple frequencies further help distinguish magnitudes and reduce periodic ambiguity. Experiments across model scales from 190M to 7B parameters show that QK-MAP achieves lower training and validation loss than both Standard attention and QK-Norm. Further analyses show that QK-MAP preserves sensitivity to input scale, assigns less attention to the first token, and produces stronger attention contrast for relevant evidence in synthetic long-context tests. These results demonstrate that query-key scale can be effectively controlled without sacrificing token-specific magnitude information.
Read the full abstract
Computer Vision
CompositionEvidence: Controlling Geometric Evidence for Cinematic Composition Reasoning with Frozen Vision-Language Models
Cinematic composition questions depend on in-frame geometry that vision-language models (VLMs) use inconsistently. External perception tools can expose this geometry, but naive prompt injection can overwrite useful visual predictions. On ShotBench composition, pre-question geometry reduces Qwen2.5-VL-7B from 49.5% to 46.3%; corrupted measurements lower it to 33.6%. CompositionEvidence derives nine reproducible features from frozen detection, depth, and line experts and supplies their shared 24-dimensional representation through two interfaces. A training-free channel appends gated neutral evidence after the question, reaching 56.2% (+6.7, exact McNemar p=0.0018) and 48.4% under corrupted evidence. A minimal-training channel maps the same representation to four soft tokens with a 7.37M-parameter projector while leaving the VLM frozen; it reaches 61.0% (+11.5). Across nine VLMs, including five commercial models queried through official provider APIs, zero-shot accuracy ranges from 43.8% to 63.0%, and the text channel changes it by +6.7 to -6.9 points; the highest baseline is among the regressions. The effect of explicit geometry depends on measurement reliability and on how the receiving model integrates it.
Read the full abstract
Game Theory
A Cooperative Game Theoretic Model for Sustainable, Smallholder Carbon Farming
Carbon farming offers a promising path to improve farmer livelihoods, mitigate climate change, and make farming sustainable. However, smallholder farmers, who dominate agriculture in developing economies, face structural barriers to participating in carbon farming due to high certification costs, limited landholdings, and limited budgets available to them. In this paper, we develop a cooperative game-theoretic framework to study how coalition formation can enable viable and profitable participation of smallholder farmers in carbon farming. Our approach explores three components: (i) Solo Portfolio Optimization (SPO), which determines an optimal portfolio of carbon farming practices that individual farmers can adopt under budget constraints; (ii) Joint Portfolio Optimization (JPO), which computes the maximum total economic surplus achievable through the formation of farmer coalitions; and (iii) surplus allocation among participating farmers in a principled way using cooperative game-theoretic solution concepts, such as Shapley value, core, least core, and nucleolus. We show that the underlying cooperative game is superadditive under an affordability condition and non-convex in general. As part of this study, we release CarbonFARM-Bench, a carbon farming dataset constructed by experts, and conduct controlled computational experiments on this dataset. We observe that solo participation in carbon farming is largely infeasible for smallholder farmers under baseline parameters, whereas coalition formation creates substantial value driven by economies of scale. In addition, we demonstrate that stability-aware allocation mechanisms are required for sustaining cooperation. Together, our results highlight that enabling smallholder carbon farming is fundamentally a challenge of institutional design, and we offer several policy recommendations to make it attractive to smallholder farmers.
Read the full abstract
Combinatorics
The cofinite regularity spectrum of $F$-irregular graphs
For a connected graph $F$, a graph $G$ is $F$-irregular if the numbers of copies of $F$ containing its vertices are pairwise distinct. We characterize which connected patterns admit regular $F$-irregular hosts of every sufficiently large degree. Stars admit no regular $F$-irregular host; for every other $F$, every sufficiently large integer $r$ is realized by a connected $r$-regular $F$-irregular graph of diameter two and order $r+o(r)$, optimal to first order. For triangles, and equivalently four-vertex paths in regular graphs, we construct such hosts of order less than $r+73r^{\log29/\log58}$ above an explicit threshold, with the constant reduced to $44$ for even $r$. Lexicographic words encode distinct rooted subgraph counts, while layered connections allow the common degree to be prescribed. Rooted-homomorphism expansions and spectral amplification establish separation for general patterns; an exact uniform-increment identity gives the quantitative triangle construction. The finite seed graphs are explicitly constructed, and their required rooted-count distinctions are verified by exact-arithmetic certificates.
Read the full abstract
Computation and Language
onóna: Twin Embeddings with Tied Unembeddings
Weight tying reduces the parameter cost of language models, but the usual tied output head is limited to a single softmax. We introduce onóna, a two-table parameterization that increases output expressiveness while preserving input-output weight sharing. The Transformer takes the average of two embedding tables as input, and the output head combines the softmax distributions defined by those same tables. A gate conditioned only on the current input token determines the mixture weight. Its values can be precomputed for inference, and the Transformer body remains unchanged. We evaluate onóna across Dense, Gated DeltaNet hybrid, mixture-of-experts, and Attention Residual architectures at 142M, 330M, and 1B scales. Under matched token budgets, onóna lowers final validation loss and external perplexity in 11 of 12 comparisons and raises the average accuracy over eight downstream tasks in 10 of 12. The one exception on loss is the 1B MoE model. The cost is one extra embedding table and one extra vocabulary projection.
Read the full abstract
Computation and Language
YRC-Bench: Benchmarking Culturally Aware Machine Translation for Historical Heritage via Tri-Level Semantic Bridging
Translating historical cultural heritage poses profound challenges for machine translation systems due to the acute semantic asymmetry across eras and cultures, where terse classical syntax frequently obscures domain-specific cultural knowledge. While Large Language Models have demonstrated remarkable prowess in generic translation benchmarks, they struggle to negotiate Culturally-Specific Items and diachronic shifts in domain-rich historical texts, such as Yellow River cultural records. To bridge this critical gap, we introduce YRC-Bench, the first human-curated, multi-parallel benchmark dedicated to historical heritage translation across Classical Chinese, Modern Chinese, and nine target foreign languages. YRC-Bench establishes a tri-level parallel corpus incorporating modern explanatory glosses as explicit semantic bridges and operationalizes translation strategies through paired conserved versus substituted CSI variants categorized by cultural translatability. Systematic evaluations of nine leading LLMs across bidirectional translation directions reveal that leveraging modern Chinese glosses - especially human-curated ones - dramatically elevates both lexical fidelity and semantic adequacy. Furthermore, multidimensional human preference audits by native experts unveil distinct cross-cultural preferences between Sinographic-tradition and non-Sinographic language groups.
Read the full abstract
Systems and Control
Balancing Confidentiality and Attack Detectability in Networked Control Systems over Lossy Wireless Channels
We study the joint protection of a networked control system (NCS) against eavesdropping and data integrity attacks over a lossy wireless channel. The eavesdropper intercepts the transmitted sensor measurements and control inputs to reconstruct the system state. In contrast, the integrity attacker in this study replaces the sensor measurements with fabricated data to drive the system to undesired operating conditions. Although these two threats are typically studied separately, a wireless NCS should be protected against both simultaneously, and a defence against one affects the other. We employ an observation-encoding scheme that renders the eavesdropper's state-estimation error covariance unbounded under packet drops, and show that transmission energy governs how long this error remains bounded. In particular, lower transmission energy improves confidentiality. For integrity protection, we combine stochastic physical watermarking with a CUSUM-based sequential change detector. The proposed attack detection scheme accounts for intermittent communication and encoded measurements. The detection delay is governed by the Kullback–Leibler divergence (KLD) between the pre- and post-attack distributions of the test signal, and we show that a higher transmission energy increases the KLD and thereby reduces the detection delay. Furthermore, we derive analytical expressions for the eavesdropper's estimation error and the worst-case expected KLD, providing guidelines for selecting transmission energy to balance confidentiality and integrity protection and characterising the privacy–security trade-off. Finally, we run numerical simulations to validate our theoretical findings and demonstrate the privacy–security trade-off in practical scenarios.
Read the full abstract
Machine Learning
Meta-Reinforcement Learning with Zero-Shot Reinforcement Learning
Meta-reinforcement learning (meta-RL) trains agents to adapt to novel tasks from test-time experience, but requires diverse training environments and reward functions. While environments are becoming cheaper to generate, it remains difficult to anticipate every objective an agent may be asked to pursue at test time. Behavior Foundation Models (BFMs) learn policies from reward-free data and support zero-shot RL (ZSRL): adaptation to new objectives at test time using a large reward-labeled dataset. Recent work has shown that BFM task inference can instead be performed online without prior task-specific data. We note that this capability positions BFMs and meta-RL as alternative approaches for online adaptation to new reward functions in a fixed environment. We then ask whether we can extend BFMs to compete with meta-RL more broadly, including in novel environments, and identify four challenges: environment identification, online data collection, test-time exploration, and learning from mixed reward supervision. We study each individually, then propose MetaBFM, a hybrid method that addresses all four. During training, it combines supervised reward-following with unsupervised learning; at test time, it can interpolate between exploration, exploitation, and ZSRL-style reward inference. We demonstrate that MetaBFM can recover the in-distribution performance of supervised meta-RL while generalizing to reward functions outside the training set. Our work suggests that BFMs and meta-RL are complementary research directions whose combined strengths allow adaptive agents to learn more diverse behaviors from a fixed set of supervised objectives.
Read the full abstract
Artificial Intelligence
Seeing Isn't Enough: Unlocking Visual Reasoning by Bridging the Reasoning-Access Gap with TIGER
Multimodal Large Language Models (MLLMs) inherit strong textual reasoning capabilities from their language backbones and text pretraining, but often fail to apply the same level of competency to visual inputs. We show that this degradation cannot be attributed to only perception failures, by evaluating models on semantically-equivalent text and rendered-text image inputs: models can accurately read text in the images and recover near-text performance when provided their own transcriptions, but just cannot reason well on the images directly. We characterize this discrepancy as a modality reasoning-access gap, where models fail to route perceived visual content into existing reasoning machinery accessible with text. To bridge this, we propose TIGER (Text-to-Image Gap-targeted training for Enhanced Reasoning), which first mines synthetic modality counterfactuals where a model succeeds on a text problem but fails when provided the corresponding rendered image, and post-trains the model on the image input via Group Relative Policy Optimization (GRPO), along with SFT and DPO variants. Importantly, although trained exclusively on synthetic rendered-text images with no multimodal annotations, TIGER results in model performance improvements across diverse real-world visual benchmarks including geometric diagrams (MathVerse), multi-disciplinary sciences (EMMA, MMMU), complex charts (ChartQAPro), and embodied spatial reasoning (RoboSpatial-Home). TIGER also complements existing state-of-the-art RLVR visual reasoning methods, yielding further performance gains. Analysis involving reasoning-subspace activation and activation patching provide support indicating that TIGER helps to align visual-to-reasoning pathways toward text reasoning states. Our implementation is available at https://github.com/minhngh/Tiger.
Read the full abstract
Statistics Theory
Removing the floor from Hoeffding and Bernstein inequalities for finite-population U-statistics
Let $C$ be a uniformly sampled $n$-element subset of a finite population $D$ of size $N$, and let $U_C$ be the complete U-statistic of order $k$ built from a bounded kernel on the $k$-subsets of $D$. We prove a product inequality for the uniform distribution on $\binom{D}{n}$ and deduce from it the comparison $ \log \mathbb{E}\left[e^{\theta(U_C-\mathbb{E}U_C)}\right] \le \frac{n}{k}\log \mathbb{E}\left[e^{\frac{k\theta}{n}(Y-\mathbb{E}Y)}\right]$, where $Y$ is a single kernel value taken at a uniformly sampled $k$-subset of $D$. This yields Hoeffding, Bennett and Bernstein inequalities for $U_C$ with effective sample size $n/k$ in place of the $\lfloor n/k\rfloor$ available so far. The gain is negligible when $n$ is much larger than $k$, but it approaches a factor of two in the exponent when $k$ is close to $n$.
Read the full abstract
Machine Learning
Dual-Granularity Learning for Regression with Continuous Noisy Labels
Learning with noisy labels has been extensively studied for classification, yet its regression counterpart remains relatively underexplored. Unlike classification, regression involves continuous-valued targets where label noise exhibits heterogeneous magnitudes across instances, which poses conflicting requirements on the model: correcting small-to-moderate deviations requires sensitivity to label values for instance-level refinement, whereas large deviations caused by corrupted or outlying annotations render exact label values unreliable, necessitating reduced reliance on exact labels and instead leveraging ordinal structure among samples. These requirements call for methods that capture both global structure and sample-specific deviations in a unified manner. To this end, we propose GRAIN, a dual-granularity framework for robust noisy regression. At the fine-grained level, we derive an explicit residual learning mechanism from a robust loss formulation, introducing instance-specific auxiliary variables to decouple label noise from predictive signals. At the coarse-grained level, we introduce a binning-based contrastive regularization that enforces an order-preserving feature structure aligned with label proximity, addressing noise beyond the capacity of local residual refinement. The two levels are coupled via a histogram-based bridging mechanism that enables bidirectional information flow: coarse-grained structural guidance establishes a well-ordered feature space that anchors residual learning, while fine-grained residual refinement corrects instance-level deviations that structural regularization alone cannot resolve. Experimental results show that GRAIN consistently improves robustness and generalization under diverse noise conditions.
Read the full abstract
Analysis of PDEs
On the domain of attraction of source–type solutions of the porous medium equations
We study the rate of convergence to Barenblatt profiles of the solutions to porous medium equations in $\R^n, n \ge 1$, by means of a certain class of Energy distances, a class of discrepancies between probability measures which have been recently widely used in problems related to AI, in particular for tasks related to machine learning. If $n \ge 3$, it is shown both that the rate of decay in Energy distance is improved with respect to known results which are based on the decay of relative entropy, and that, with respect to this distance, any initial value of bounded mass belongs to the domain of attraction of the source–type solution. Similar results are shown to hold for the fast diffusion case.
Read the full abstract
Astrophysics of Galaxies
The hierarchical structure of the Milky Way I. A model-agnostic clustering of 1 billion stars in 5D
The Milky Way exhibits a hierarchical structure shaped by mergers, accretion, and internal dynamics. Gaia Data Release 3 (DR3) enables a global view of this structure, but requires scalable, model-agnostic methods to extract multi-scale clustering patterns. We aim to characterise the hierarchical clustering structure of the Milky Way in a model-agnostic way and to assess how the resulting structures relate to existing catalogues. We apply the hierarchical density-based clustering algorithm AstroLink to a locally magnitude-complete subsample of over one billion Gaia DR3 stars with five-dimensional astrometric information. The data are corrected for selection effects and transformed using a distance contraction and feature scaling. We compare the resulting cluster hierarchy to catalogues of open clusters, globular clusters, dwarf galaxies, and stellar streams using custom availability-aware metrics and a bi-directional best-match framework. At a significance threshold of $S = 3.8\sigma$, AstroLink identifies 7410 statistically significant clusters spanning a wide range of morphologies. Compact systems are often recovered as reciprocal matches with many also exhibiting internal substructure, while extended structures are typically fragmented or partially recovered. A substantial fraction of clusters have no counterpart in the comparison catalogues, these may be new structures but additional physical checks are needed to verify this. A model-agnostic, hierarchical clustering approach can recover the multi-scale structure of the Milky Way at Gaia DR3 scale and provides a unified framework for comparing heterogeneous catalogues and identifying candidate structures.
Read the full abstract
Robotics
HSGNav: Hypothesis Scene Graphs for Multi-Scale Lifelong Navigation under Object Relocation
In real homes, assistive robots must repeatedly locate targets from large furniture to tiny tabletop items, as people move objects between household surfaces. Uniformly instantiating detections at different scales as 3D objects can be brittle for small on-surface items; without explicit object-state validation, stale entries may persist after scene changes. Our key insight is to separate persistent 3D furniture entities from revisable surface-object hypotheses, retaining their visual evidence without requiring independent object geometry. We formalize this design as the Hypothesis Scene Graph (HSG), which preserves small-object evidence and supports local revision without rebuilding the stable entity topology. Building on HSG, HSGNav uses this separation for multi-scale lifelong navigation: retrieval selects entity or hypothesis evidence according to target scale, exploration acquires additional surface observations, and verification confirms target presence or invalidates contradicted hypothesis records. Because existing lifelong benchmarks such as GOAT-Bench rarely require mid-episode memory invalidation or multi-scale search, we introduce FAROS-Bench (Find And Re-find Objects across Scales) for mid-episode relocation of surface objects and multi-scale find goals. Experiments show that HSGNav performs competitively on GOAT-Bench while substantially outperforming prior methods on FAROS-Bench.
Read the full abstract
Soft Condensed Matter
Pinching an elastic potato chip
Thin shells deform by transmitting stress across their surface, yet how a shell of general shape carries a localized load remains unclear. Combining experiments, simulations, and theory on pinched saddle-shaped elastic shells, we show that the asymptotic lines, curves of zero normal curvature, dictate both deformation and stiffness. Isometry turns the shell mid-surface into a medium obeying a wave equation in space. Material points lit by the pinch are lifted, while those in its shadow sink. Finite thickness blurs the lit–shadow boundary, and the same rule maps where a clamped shell is rigid. Our findings indicate that light and shadow offer a unified framework for understanding mechanical stress propagation along shell structures.
Read the full abstract
Software Engineering
What extraction cannot slim, and whether function injection can: a pre-registered study of 580 Java repositories
A method of 26 code lines or more reads as a Long Method, and Extract Method is the usual fix. Some long methods have no single-exit fix: a return cannot leave an extracted piece, a piece hands back one value, and some methods are knotted so that every cut is blocked. Function injection, as the OfficeFloor framework does it, cuts differently: each piece becomes a function the framework injects and sequences, an early return ends a function, and a function hands on several values. The framework's author pre-registered a test of whether it takes apart what extraction cannot. A stratified random sample of 580 open-source Java repositories, each pinned before AI coding assistants were in general use, was scanned by released tools whose cost models were fixed before the sample was drawn: 3,026,902 methods, 168,466 of them long. Single-exit extraction cannot slim 22.9% of the long methods. The registered hypothesis, that at least a quarter of those stuck methods are ones only function injection takes apart, is refuted: 22.2% (95% CI 20.3 to 24.4). Nearly half of the stuck methods are early-return methods the plan counts as conventionally fixable, and the split takes three in four of those apart as they stand. Where extraction would do, the split chooses injection and costs less for 29.0% against extraction wrapped as one function and 17.3% against plain extraction, almost always as a straight sequence of functions. An exploratory reading of the quarter of the stuck methods that neither takes apart finds a third of it tables and literals, long for their data rather than their logic, and most of the rest blocked by a jump inside a loop, by a nested statement too large to cut, or by values too many for any cut. Three registered studies are proposed: the classes as rules in the tools, a refinement per class, and the same measures on code written with AI assistants.
Read the full abstract
Quantum Physics
Compact Quantum Channel Attention for Static ASL Recognition: A Controlled Study with Parameter-Matched Baselines
Hybrid quantum-classical models often report good accuracy on image tasks, but it is rarely clear whether the quantum part is responsible. This paper proposes QCA-DenseNet, a static American Sign Language (ASL) recognition pipeline in which an eight-qubit variational circuit acts as a channel-attention gate on DenseNet121 features. The quantum channel attention (QCA) gate has 17,464 parameters, 7.51 times fewer than a squeeze-and-excitation (SE) block. The pipeline is tested on 1,815 MU HandImages images by holding out one of five signers at a time, and a frozen-backbone ablation uses small parts of the Kaggle ASL Alphabet data. On unseen signers, QCA-DenseNet reaches 94.50% mean accuracy, above SE (93.72%) and within 0.28 points of the best model. A matched classical MLP and a model without attention perform about the same, and no difference is significant. QCA gives a much smaller attention gate without a detectable loss in accuracy, but no quantum advantage.