arXiv:2607.07391v1 Announce Type: new Abstract: Mathematical reasoning benchmarks typically provide all facts needed to solve each problem, while interactive benchmarks often mix reasoning with tools, retrieval, and long-horizon dialogue. We introduce MIRA-Math, a benchmark for a narrower diagnostic capability: solving mathematical problems whose full latent state has a unique answer, but whose solver-facing view is missing exactly one necessary atomic fact. The solver must request the missing information in natural language under a strict budget and then integrate the returned fact into an exact final answer. A fixed constrained LLM responder sees only the dataset-provided atomic fact and must either offer the quoted fact when the request matches it, or decline otherwise. Thus, instance generation, typed hint specifications, validation, and final-answer verification are deterministic, while request metrics are measured under a fixed LLM-mediated responder channel. MIRA-Math contains 2{,}310 generated instances from 22 typed mathematical families spanning algebra, probability, linear systems, discrete structures, signal processing, Markov chains, circuits, interpolation, and numerical boundary-value problems. Experiments across frontier and small models show that request success and final-answer accuracy are separable: models may ask for the right fact yet fail the downstream computation, or fail before obtaining the canonical hint. We release generators, verifiers, prompts, run metadata, and dataset documentation to support reproducible evaluation of minimal information requesting in mathematical reasoning.
Science Journals
arXiv:2510.26360v3 Announce Type: replace Abstract: Materials that dissipate energy efficiently under high-speed impacts, from micrometeoroid strikes on spacecraft to ballistic penetration in protective systems, are essential for maintaining structural integrity in extreme environments. Yet, despite decades of study, predicting and comparing impact performance across materials, geometries, and length scales remains challenging because conventional projectile-impact models often rely on conservation-based or empirically partitioned descriptions that assume the projectile-target interaction is a closed system. Here, we relax this assumption and directly observe the momentum and energy transferred out of the projectile during impact. We find that the momentum transferred to the target consistently reaches its maximum at the ballistic-limit velocity, demonstrated through a coordinated suite of micro-projectile impact experiments spanning varied projectile diameters, target thicknesses, and impact velocities, and further supported by targeted macroscale tests. This behavior is reinforced across a broad range of independent studies encompassing metals, polymers, composites, sandwich panels, and reinforced concrete, with thicknesses ranging from nanometers to hundreds of millimeters and projectiles of spherical, blunt, ogive, and conical shape, under both normal and oblique impacts. Together, these observations reveal a consistent impact behavior across all available data: maximum momentum transfer occurs at the ballistic limit. Extending this bound into the energy absorption landscape addresses an entrenched misconception in the field by revealing that specific energy absorption inherently inflates the performance of thinner targets due to geometric normalization, rather than reflecting genuine material enhancement.
arXiv:2511.01043v2 Announce Type: replace Abstract: Large Language Models (LLMs) are increasingly used in software engineering tasks, especially code repair. However, developers often struggle to interpret model outputs, limiting effective human--AI teaming, where humans and AI work toward a shared objective. Prior work mainly optimizes generated code, giving less attention to natural-language feedback that supports comprehension and iterative improvement. We present \textsc{DPO-f+}, a framework that aligns code-repair feedback with the needs of different developer groups, including novices and proficient developers. It (1) defines feedback-alignment metrics across seven fixed dimensions with task-specific descriptions; (2) automatically constructs pairwise preference datasets from code-repair tasks; (3) fine-tunes models using Direct Preference Optimization (DPO) augmented with a reward model; and (4) provides an automated protocol for evaluating feedback quality. Empirically, \textsc{DPO-f+} outperforms both the baseline and standard DPO in feedback accuracy and overall alignment. On novice programming tasks, \textsc{DPO-f+} improves Pass@1 by 5.71 percentage points (pp) over the baseline and 3.30 pp over DPO. On \textsc{SWE-Bench}, it improves issue-resolution rate by 1.67 pp over DPO and 4.67 pp over the baseline. It also improves feedback alignment by both LLM judges and a human study with 200 developers: beginners preferred \textsc{DPO-f+} in 71.5% of comparisons, with overall preference above chance ($p=0.0057$). By better aligning feedback with developer needs, \textsc{DPO-f+} turns LLM assistance from a one-shot output into a collaborative sense-making workflow, enhancing code comprehension and human--AI teaming in software engineering.
arXiv:2511.06667v2 Announce Type: replace Abstract: With the explosive growth of rigid-body simulators, policy learning in simulation has become the de facto standard for most rigid morphologies. In contrast, soft robotic simulation frameworks remain scarce and are seldom adopted by the soft robotics community. This gap stems partly from the lack of easy-to-use, general-purpose frameworks and partly from the high computational cost of accurately simulating continuum mechanics, which often renders policy learning infeasible. In this work, we demonstrate that rapid soft robot policy learning is indeed achievable via implicit time-stepping. Our simulator of choice, DisMech, is a general-purpose, fully implicit soft-body simulator capable of handling both soft dynamics and frictional contact. We further introduce delta natural curvature control, a method analogous to delta joint position control in rigid manipulators, providing an intuitive and effective means of enacting control for soft robot learning. To highlight the benefits of implicit time-stepping and delta curvature control, we conduct extensive comparisons across four diverse soft manipulator tasks against one of the most widely used soft-body frameworks, Elastica. With implicit time-stepping, parallel stepping of 500 environments achieves up to 6x faster speeds for non-contact cases and up to 40x faster for contact-rich scenarios. Finally, a comprehensive sim-to-sim gap evaluation--training policies in one simulator and evaluating them in another--demonstrates that implicit time-stepping provides a rare free lunch: dramatic speedups achieved without sacrificing accuracy.
arXiv:2604.19238v2 Announce Type: replace Abstract: Real-world image super-resolution (Real-SR) has been revolutionized by leveraging the powerful generative priors from Diffusion Models (DMs) and Flow Matching (FM). However, existing one-step methods typically replace Gaussian noise with degraded low-resolution (LR) latents at initialization, introducing a substantial distribution shift that further leads to trajectory deviation and prior collapse under extreme acceleration. To overcome these limitations, we propose Allo{SR}$^2$, a novel FM-based framework that rectifies one-step SR flows via allomorphic generative flows to maintain high-fidelity generative realism. Specifically, we utilize SNR-Guided Trajectory Initialization to identify a statistically aligned intermediate state along the pre-trained path to integrate LR representations into the generative flow. To ensure a stable, low-curvature path for one-step inference, we propose Flow-Anchored Trajectory Consistency (FATC), which explicitly regularizes the velocity field of the underlying probability flow. Furthermore, we develop Allomorphic Trajectory Matching (ATM), a self-adversarial distillation strategy that jointly models the SR flow and the generative flow within a unified velocity field, enabling one-step Real-SR while preserving the generative prior. Extensive experiments on both synthetic and real-world benchmarks demonstrate that Allo{SR}$^2$ achieves state-of-the-art performance in one-step Real-SR, offering a superior balance between fidelity and realism while maintaining extreme efficiency.
arXiv:2511.23347v2 Announce Type: replace Abstract: An associative memory (AM) enables cue-response recall, and it has recently been recognized as a key mechanism underlying modern neural architectures such as Transformers. In this work, we introduce the concept of distributed dynamic associative memory (DDAM), which extends classical AM to settings with multiple agents and time-varying data streams. In DDAM, each agent maintains a local AM that must not only store its own associations but also selectively memorize information from other agents based on a specified interest matrix. To address this problem, we propose a novel tree-based distributed online gradient descent algorithm, termed DDAM-TOGD, which enables each agent to update its memory on the fly via inter-agent communication over designated routing trees. We derive rigorous performance guarantees for DDAM-TOGD, proving sublinear static regret in stationary environments and a path-length dependent dynamic regret bound in non-stationary environments. These theoretical results provide insights into how communication delays and network structure impact performance. Building on the regret analysis, we further introduce a combinatorial tree design strategy that optimizes the routing trees to minimize communication delays, thereby improving regret bounds. Numerical experiments demonstrate that the proposed DDAM-TOGD framework achieves superior accuracy and robustness compared to representative online learning baselines such as consensus-based distributed optimization, confirming the benefits of the proposed approach in dynamic, distributed environments.
arXiv:2512.02076v2 Announce Type: replace Abstract: We propose FDRMFL, a task-driven multimodal feature extraction framework for federated regression under non-IID data distributions. Extracting predictive features from high-dimensional multimodal inputs is particularly challenging in this setting: data cannot leave each client, local samples are scarce and heterogeneously distributed, and unsupervised dimensionality reduction discards task-relevant information while federated training introduces representation drift across communication rounds. FDRMFL addresses these challenges through a unified four-term local objective: MSE prediction loss, a correlation-based mutual information surrogate that preserves dependence between the fused representation and the continuous target, a symmetric KL penalty that aligns cross-modal latent distributions before fusion, and an InfoNCE-style contrastive loss that anchors local representations to the global consensus. Experiments on three synthetic and two real-world near-infrared spectroscopy datasets under non-IID federated partitions, with comprehensive ablation and sensitivity analyses, demonstrate that each component contributes to the framework's effectiveness. FDRMFL reduces mean MSE by 33.8% relative to the best traditional baseline (PCA) and by 43.0% relative to VAE in simulation, and attains the lowest overall mean MSE among six federated algorithms including FedAvg, FedProx, MOON, SCAFFOLD, and FedBN.
arXiv:2607.07131v1 Announce Type: new Abstract: Modeling the quantum dynamics of plasmonic excitations -- collective oscillations of free electrons interacting with light -- remains a significant theoretical challenge, particularly due to the need to accurately describe their quantum nature and the role of non-radiative decay channels. At the same time, a reliable theoretical framework is essential for advancing applications ranging from materials design to the development of new quantum optical platforms for quantum technologies. In this work, we address these challenges by introducing a Hermitian formalism based on the linear vibronic coupling (LVC) model for the description of plasmonic excitations in metallic nanostructures. This is parameterized through first-principles calculations -- including but not limited to, the full DFT ground state with tight-binding excited states -- and machine learning techniques using a newly implemented automated platform named Python Plasmonic Cavity (PyPC). The effectiveness of this workflow is demonstrated by successfully reproducing the experimental absorption spectra and vibronic broadening of plasmonic silver nanoparticles containing more than a hundred atoms. Additionally, the population dynamics of plasmonic states are investigated, showing that the LVC model accurately predicts ultrafast lifetimes for bright states and effectively captures the dynamics of dark states.
arXiv:2607.06869v1 Announce Type: cross Abstract: We show that for every $k \in \mathbb{N}$, every graph $G$ contains $k$ vertex-disjoint cycles of different lengths, or there exists a set $X \subseteq V(G)$ with $|X| \in \mathcal{O}(k^6\mathsf{polylog}(k))$ such that $G-X$ has at most $k-1$ cycle lengths. We also prove analogous results for facial lengths of embedded graphs. Let $G$ be a graph with a closed 2-cell embedding $\psi$ on a surface $\Sigma$ of Euler genus $g$, let $c$ be a colouring of the faces $\mathcal{F}(\psi)$ of $\psi$, and let $R(G,\psi)$ be the radial graph of $(G, \psi)$. Then there exist $k$ faces $F_1, \ldots , F_k \in \mathcal{F}(\psi)$ that are given pairwise distinct colours by $c$ and are pairwise at distance at least $d$ in $\psi$, or there exists a set $X \subseteq V(G)$ of order at most $\mathcal{O}(k^2dg)$ such that $|\{ c(F) \mid F \in \mathcal{F}(\psi) \text{ and } V(F) \cap \bigcup_{x \in X} N^d_{R(G,\psi)}(x) = \emptyset \}| \leq k(k+2)$. Finally, using a result from additive combinatorics, we show that there are subdivided ladders with only a small number of cycle lengths. This suggests that it may be difficult to improve our bounds.
arXiv:2606.08336v2 Announce Type: replace Abstract: While multimodal integration significantly improves computer vision models, deploying them incurs prohibitive inference costs and requires scarce, perfectly paired datasets. Recent methods address this data bottleneck by synthesizing missing modalities via generative AI, yet they introduce a severe inefficiency: the Decode-Encode Loop. Specifically, information-rich generative latents are decoded into noisy raw signals, forcing the downstream classifier to waste capacity re-encoding them. To bypass this bottleneck, we propose Direct Latent Augmentation (DLA), utilizing undecoded generative latents directly as privileged information. Furthermore, to transfer this dense knowledge to a purely visual student, we introduce Multilayer Explicit Simulated Synesthesia (MESSy). Instead of enforcing rigid representation matching, which forces the student to distort its native visual features to accommodate complex multimodal topologies, MESSy uses a predictive objective to safely internalize these physical priors. Empirical results demonstrate that our framework significantly outperforms raw data augmentation and traditional distillation. Ultimately, our approach yields highly accurate unimodal students with "synesthetic" latent structures that are inherently aligned with modalities they have never directly observed.
arXiv:2606.11852v2 Announce Type: replace Abstract: For a set $X$ of integer points, the relaxation complexity $\operatorname{rc}(X)$ is the smallest number of facets of any polyhedron $P$ such that $P \cap \mathbb{Z}^d = X$. In this paper, we focus on the case where $X$ is the discrete standard simplex $\Delta_d = \{\mathbf{0}, \mathbf{e}_1, \dots, \mathbf{e}_d\}$. We show that $\operatorname{rc}(\Delta_d) = O(\log d)$ by an explicit, elementary construction. This improves upon the previously best-known upper bound $\operatorname{rc}(\Delta_d) = O(d / \sqrt{\log d})$ due to Aprile, Averkov, Di Summa, and Hojny (2024) and matches an asymptotic lower bound by Averkov and Schymura (2022).
Miter-Aware LUT Mapping: Aligning Structure and Solvability for Efficient Logic Equivalence Checking
arXiv:2607.07164v1 Announce Type: new Abstract: Logic Equivalence Checking (LEC), a fundamental hardware verification task, is often bottlenecked by synthesis-induced structural perturbations and XOR-dense regions that degrade SAT solver performance. We contend that the modeling of the miter is as critical as the SAT solver itself. To this end, we introduce a miter-aware mapping framework that strategically formulates the problem before solving. By constructing a LUT-based miter -- instead of a traditional, flat netlist -- our approach preserves critical structural correspondence between the two designs while making high-level logic relations explicit. Our framework uniquely integrates three techniques: equivalence-preserving mapping to structurally align the two circuits, Gaussian-guided XOR modeling to algebraically simplify dense arithmetic, and solver-oriented LUT selection to generate a representation optimized for efficient SAT reasoning. Evaluated on comprehensive datasets, our method achieves up to a \textbf{92.1\%} reduction across state-of-the-art SAT solvers. This demonstrates that a solver-aware modeling paradigm, which unifies structural mapping with SAT reasoning, can fundamentally enhance LEC efficiency.
arXiv:2607.07514v1 Announce Type: new Abstract: The simulation of electric machines plays a significant role in the design of efficient and competitive products. Faster simulations reduce computational costs, such as CPU hours, and shorten development cycles, thereby enabling faster design iterations and ultimately accelerating time-to-market. In this work, we analyze the dominant computational bottlenecks and demonstrate how targeted acceleration measures can significantly reduce the overall runtime of 2D and 3D finite element simulations of electric machines in an industrial environment.
arXiv:2607.07517v1 Announce Type: new Abstract: We study bimatrix two-player games and investigate the last-iterate convergence and stability of equilibria for the iterates generated by the optimistic exponential weights method. In contrast to prior work, we allow the step sizes $\eta_x$ and $\eta_y$ to differ. Our first main result establishes, under the assumption that the set of fixed points is finite, a sufficient condition for global last-iterate convergence in the special case of zero-sum games, which constrains only the product $\eta_x\eta_y$ of the step sizes. This condition is practically relevant and partially explains empirically observed behavior. Our second main result provides an almost-tight threshold for asymptotic stability and instability, again in terms of products of the step sizes, for general bimatrix games. This result is primarily of theoretical interest. We derive several known results and practically relevant step size bounds for special cases and illustrate our results by experiments.
arXiv:2604.07935v2 Announce Type: replace Abstract: The Hardware Lottery posits that research directions are dictated by available silicon compute platforms. We identify a derivative phenomenon, the Hyperscale Lottery, where model architectures are optimized for cloud throughput at the expense of algorithmic efficiency. While State-Space Models (SSMs) such as Mamba were lauded for their linear complexity, ideal for edge intelligence, their evolution from Mamba-1 to Mamba-3 reveals a systematic divergence from edge-native efficiency. We demonstrate that Mamba-3's architectural changes, designed to saturate hyperscale GPUs, impose a significant edge penalty: a 28% latency increase at 880M parameters, worsening to 48% for 15M-parameter models. We argue for decoupling cloud-scale saturation strategies from core architectural design to preserve the viability of single-user, real-time edge intelligence.
arXiv:2607.07195v1 Announce Type: new Abstract: Perceptual quality enhancement of severely compressed videos remains challenging due to complex artifact patterns and substantial information loss. Recent diffusion models have demonstrated strong generative capability for visual restoration, but directly applying them to compressed video often ignores compression degradation characteristics and may introduce structure-inconsistent hallucinations. To address this issue, this paper presents a diffusion-based compressed video enhancement method, named DiffCVE. Coding Prior-enhanced Dual Conditioning (CPDC) branches are designed to jointly model compressed video and coding prior conditions, where coding priors including residuals and motion vectors provide complementary structural and motion guidance during the diffusion denoising process. To make the diffusion process aware of compression severity, a Compression Degradation Semantic Prompting (CDSP) mechanism is introduced to leverage QP-conditioned textual prompts together with LoRA fine-tuning. In addition, a Coding Prior-guided Weighted Fusion (CPWF) module is incorporated into the VAE decoder to fuse VAE encoder and coding prior encoder features with QP-predicted weights. Extensive experiments demonstrate the effectiveness of the proposed method in improving perceptual quality, especially under severe compression settings. The project page with enhanced video demonstrations is available at https://wqmaker.github.io/projects/DiffCVE/.
arXiv:2601.16529v4 Announce Type: replace Abstract: Large language models (LLMs) deployed in clinical decision support may acquiesce to patient requests for care that conflicts with evidence-based guidelines. We developed SycoEval-EM, a multi-agent simulation framework to evaluate LLM robustness to adversarial patient persuasion in emergency medicine. Across 19 contemporary LLMs and 1,425 simulated clinical encounters spanning three Choosing Wisely scenarios, acquiescence rates ranged from 0% to 100%, revealing a bimodal distribution. Seven models maintained near-perfect guideline adherence, while six acquiesced in the majority of encounters. Vulnerability varied substantially across clinical scenarios. Acquiescence was highest for CT imaging requests, intermediate for antibiotic prescriptions for sinusitis, and lowest for opioid prescriptions for acute back pain. Model scale, recency, and performance on static medical benchmarks did not consistently predict robustness. All five persuasion tactics produced similar acquiescence rates, with no statistically significant differences after correction for multiple comparisons, suggesting a generalized susceptibility rather than tactic-specific weaknesses. LLM-as-judge evaluation was validated against two independent physician raters across 95 matched conversations and demonstrated near-perfect agreement for the primary outcome of acquiescence (Cohens kappa = 0.957). These findings indicate that static medical benchmarks are insufficient to predict safety performance under sustained social pressure and support incorporating multi-turn adversarial testing into clinical AI evaluation. Notably, two models achieved perfect guideline adherence across all encounters, demonstrating that robustness to patient pressure is attainable without sacrificing effective clinical communication.
arXiv:2601.13563v5 Announce Type: replace Abstract: In current Mixture of Experts (MoE) architectures, linear memory scaling is present, the memory grows as the number of experts increases. $N$ independent expert weight matrices require $\mathcal{O}(N \cdot d^2)$ memory which exceeds the memory budget of edge devices. Current compression methods like quantization, pruning, and low-rank factorization reduce constant factors, but the scaling bottleneck is still unresolved. We introduce ButterflyMoE, a method which treats experts not as independent matrices but as geometric reorientations of a shared quantized substrate. Diversity amongst the experts arises from viewing different angles of the shared capacity and not from redundant storage. Learned rotations are applied to a shared ternary prototype. With this, each expert yields $\mathcal{O}(d^2 + N \cdot d \log d)$ memory-reducing per-expert cost from $\mathcal{O}(d^2)$ to $\mathcal{O}(d \log d)$. The key insight is that training these rotations with quantization reduces activation outliers and stabilizes extreme low-bit training where other static methods collapse. Across language modeling benchmarks, ButterflyMoE achieves 80$\times$ memory reduction at 8 experts with a highly favorable memory-accuracy tradeoff.At this 80x compression ButterflyMoE outperforms an equal memory dense baseline, showing that orbital parameterization extracts fundamentally more utility per byte. When scaled up to 256 experts, ButterflyMoE asymptotically compresses the memory by 150 $\times$. ButterflyMoE reduces the constant factor of linear scaling with compression ratio growing with the expert count.
arXiv:2601.16946v2 Announce Type: replace Abstract: Large language models (LLMs) are increasingly used for text analysis tasks, such as named entity recognition or error detection. Unlike encoder-based models, however, generative architectures lack an explicit mechanism to refer to specific parts of their input. This leads to a variety of ad-hoc prompting strategies for span labeling, often with inconsistent results. In this paper, we categorize these strategies into three families: tagging the input text, indexing numerical positions of spans, and matching span content. To address the limitations of content matching, we introduce LogitMatch, a new constrained decoding method that forces the model's output to align with valid input spans. We evaluate all methods across four diverse tasks. We find that while tagging remains a robust baseline, LogitMatch improves upon competitive matching-based methods by eliminating span matching issues and outperforms other strategies in some setups.
arXiv:2606.22568v4 Announce Type: replace Abstract: Training image generation foundation models consumes substantial resources. Previous methods have attempted to leverage semantic guidance to accelerate the training process, yet their experiments were only conducted on simple datasets such as ImageNet, at low resolutions, and with small-scale models. In this paper, we propose SeFi-Image, a text-to-image foundation model built upon semantic-first diffusion, a novel latent diffusion modeling paradigm. We instantiate SeFi-Image at three model scales, 1B, 2B, and 5B parameters, enabling systematic study of scaling behavior and flexible deployment under varying compute budgets. Notably, our largest 5B model was trained with merely 125K A800 GPU hours, corresponding to roughly 10-20% of the training compute used by Z-Image. However, it achieves results comparable to or even superior to Qwen-Image and Z-Image. Despite this modest training compute, SeFi-Image achieves strong performance on a wide range of benchmarks, including GenEval, DPG, LongTextBench, OneIG, and CVTG-2K. Moreover, we provide DMD2-distilled few-step turbo variants for each model scale to accommodate diverse hardware constraints and latency requirements. We publicly release our code, weights and hope this work offers the community useful insights into semantic-guided diffusion modeling for T2I generation, while also providing practical and readily deployable model options.
arXiv:2607.05876v2 Announce Type: replace Abstract: LLM serving optimization typically benchmarks many configurations and reaches for heavy profilers when latency targets are missed. We argue for the reverse discipline: estimation is the analytical layer of profiling -- without it, optimization degenerates to grid search. Floor First is a residual-driven triage workflow. Each decode step is modeled as a five-dimensional resource vector (HBM bytes, FLOPs, network bytes, network messages, KV capacity); summing within a resource and maximizing across resources gives an optimistic floor, the plain sum a pessimistic one. Where a measurement lands inside this [max, sum] interval reads out overlap quality before any profiler is opened, and profilers escalate only on residuals above a stated threshold. Deployment alternatives are compared by wall ordering -- which resource wall binds first as load grows -- rather than by point benchmarks. The account is compositional: new attention or state-space variants enter by declaring one module, and the workflow ships as a zero-dependency calculator plus an agent skill that enforces the discipline in agentic optimization loops. As a case study we analyze a DeepSeek-V3.2-style 671B MoE/MLA model on 16 NVIDIA H20 GPUs, whose ridge point of ~74 FLOP/byte (vs ~590 for H100) makes it an extreme decode-oriented part. The floors show TP16 decoding is KV-capacity-limited to ~70 concurrent 8K requests; sparse attention removes the KV-bandwidth term but not the capacity wall; an EP16+DP-attention layout accepts slightly worse same-batch weight traffic for an order-of-magnitude higher capacity wall (~644) -- while single-stream latency favors TP by 2.4x. The layout judgment is thus a computable function of the operating point, explaining why production deployments on identical hardware have shipped opposite attention layouts.
arXiv:2606.25075v3 Announce Type: replace Abstract: Computing the viscous Stokes flow around a shape requires solving a boundary-integral equation, and for a new shape the solve must begin from scratch. Learned operators promise to spread this cost across shapes, but it is not clear what such an operator retains of the solver it replaces, or what determines whether it transfers to shapes it was not trained on. We make both questions answerable by choosing a problem which is exactly solvable except for a single term: a second-kind boundary-integral problem is solved exactly using a kernel-independent fast summation, while the boundary correction, which has no closed form, is learned. The solver's guarantees carry over unchanged: exactness on the closed-form part, $O(N)$ scaling, $SO(3)$-equivariance to machine precision, and an $O(N)$ differentiable adjoint. We then make precise what the learning contributes. It does not contribute to accuracy, differentiability, or scaling $O(N)$, all of which are provided by the solver. Learning contributes a one-time cost in that the forward map is trained once and then evaluated on a new shape in a single pass rather than resolved. Measured against baselines, the learned map is $5$ to $16\times$ more data-efficient than a black-box DeepONet and maintains a $\sim\!2.5\times$ lower in-distribution error than a geometry-aware operator, although that operator is stronger in the low-data limit and the learned map is less reliable out of distribution. We trace this fragility to the global parameterization and reduce it with a local equivariant kernel. The exactly-solvable setting yields clarity about the mechanism: geometric generalization is governed by invariance and coverage, not by conditioning or by capacity.
arXiv:2606.29744v2 Announce Type: replace Abstract: Accurate coronary Digital Subtraction Angiography (DSA) vessel segmentation is essential for computer-aided diagnosis and treatment planning of coronary artery disease (CAD). However, thin low-contrast vessels, background interference, and severe vessel-background class imbalance make reliable segmentation of weak distal branches and vessel boundaries challenging. Existing methods struggle to balance global contextual reasoning with preservation of weak vessels, vessel continuity, and fine boundaries. To address these limitations, we propose HTC-SGA Former, a lightweight hybrid Transformer-CNN framework for coronary DSA vessel segmentation. It employs a CNN encoder for local vessel morphology extraction and a Transformer decoder for contextual feature modeling. A Multi-Scale Global-Local Window Attention (MS-GLWA) block performs efficient global-local contextual modeling, while a Self-Guided Feature Attention (SGFA) module enhances weak-vessel responses. In addition, a Boundary-Weighted Adaptive Compound Loss (BWACL) emphasizes thin-vessel boundaries and adaptively balances vessel recovery and boundary refinement. Experiments on private right and left coronary artery DSA subsets show that HTC-SGA Former outperforms 14 state-of-the-art segmentation methods while maintaining a compact architecture with only 0.81M parameters. BWACL also improves performance over binary cross-entropy and Dice losses across four encoder-decoder architectures, demonstrating strong cross-backbone applicability. HTC-SGA Former improves thin-vessel recovery, vessel continuity, and boundary localization through complementary global-local contextual modeling, vessel-focused refinement, and adaptive optimization, supporting reliable and computationally efficient coronary vessel analysis for future computer-assisted cardiovascular interventions.
arXiv:2607.06076v2 Announce Type: replace Abstract: Computerized gait analysis (CGA) serves as an essential diagnostic tool for evaluating neuromuscular, musculoskeletal, and neurological disorders in children, from cerebral palsy to muscular dystrophy. By enabling objective and comprehensive gait analysis, CGA supports timely clinical interventions that can significantly improve pediatric mobility outcomes and quality of life. Yet pediatric gait analysis introduces unique design considerations often underexplored in existing CGA research, as children's ongoing development shapes assessment requirements. To understand how CGA technologies can be designed for pediatric care, we conducted a qualitative study with 12 pediatric clinicians and one system designer who routinely work with CGA. Participants identified child-specific challenges including managing heightened sensory sensitivities to wearable devices, accommodating body proportions in sensor placement and calibration, and maintaining patient engagement during data collection. Clinicians also articulated needs for workflow adaptations and expressed interest in extending gait analysis beyond controlled laboratory settings into naturalistic environments such as playgrounds and schools, where children's authentic movement patterns emerge. Drawing from these clinician perspectives, we present design recommendations for pediatric-centered CGA that address sensing modalities suitable for sensory-sensitive children and approaches for capturing gait data across diverse care environments. Our findings contribute to the broader challenge of adapting clinical technologies to meet the distinct needs of pediatric populations.
arXiv:2607.06149v2 Announce Type: replace Abstract: There has been growing public interest in understanding personality traits and emotional characteristics, as such knowledge helps individuals better accept themselves and manage negative emotions. While professional personality scales remain the standard tool for assessment, they are often perceived as tedious or inaccessible to the general public. AI-driven systems can make assessments more accessible, but it is difficult to balance user engagement with predictive consistency in existing works. We tackle this challenge by introducing BlossomPsy, a user-friendly AI-driven MBTI assessment system. MBTI, a widely recognized but psychometrically debated personality framework, serves as the foundation for many recent systems. BlossomPsy integrates multi-turn dialogue and photo-based questions to enhance user engagement while supporting confidence-aware predictions. By combining deep learning, multi-armed bandit algorithms, and control theory, the system dynamically adapts to users' responses. In particular, photo-based questions are designed to increase interactivity and provide additional user information, thereby improving prediction confidence. Experiments involving both human volunteers and large language models (LLMs) provide preliminary evidence that BlossomPsy can produce stable predictions, with higher reported user satisfaction compared to MBTI-M (Chinese version), while maintaining comparable consistency with the reference scale.