arXiv:2607.14117v1 Announce Type: new
Abstract: Cross-version differencing of scientific documents is essential in scholarly publishing and technical documentation, but remains challenging because scientific documents are page-structured artifacts containing heterogeneous elements such as text, tables, formulas, figures, and layout cues. Existing text-sequence-based methods often lose layout and structural information, while image-based methods lack semantic interpretability and are sensitive to rendering variation. To address these limitations, this paper proposes a layout-aware heterogeneous element-aware framework for scientific document differencing. The framework decomposes document versions into semantically typed elements, establishes cross-version correspondence through an alignment-first mechanism that jointly models spatial, content, and structural compatibility, and performs type-aware difference reasoning over aligned element pairs. It supports unified change detection, localization, structure-awareness analysis, and alignment/matching evaluation across text, tables, formulas, and figures. Experiments on real-world scientific PDF data from journal production proofreading workflows show that the proposed framework consistently outperforms element-specific baselines. It achieves detection F1 scores of 0.903, 0.855, 0.862, and 0.845 for text, tables, formulas, and figures, respectively, with further improvements in localization, structure awareness, and matching quality. Ablation and sensitivity analyses confirm the effectiveness of cross-version alignment, type-specific representations, structure-aware reasoning, and compatibility-weight design. These results demonstrate that heterogeneous element-aware differencing provides a robust and interpretable solution for scientific document comparison in realistic editorial production scenarios.
Science Journals
arXiv:2607.14324v1 Announce Type: new
Abstract: Incremental deployment of hollow-core fiber (HCF) in single-mode-fiber (SMF) networks introduces a routing tradeoff: reducing HCF-SMF transitions can improve
physical-layer feasibility, but overly transition-averse routing incurs harmful path detours. We study this tradeoff using a common event-driven simulator
that compares six protected routing schemes spanning fiber-blind, generalized signal-to-noise ratio (GSNR)-aware, and explicitly transition-aware designs on
hybrid HCF/SMF topologies. The model includes a per-transition GSNR penalty and an exploratory splice-failure availability term. Across six reference
topologies, five HCF deployment fractions, and dynamic loads at 300 Erlang, the strongest transition minimizers, transition-penalty-aware routing (TPAR) and
the GSNR/fiber-transition joint scheme (GFJ), halve the mean transition count at a 20-25% carried-traffic penalty. Among the intermediate designs, GSNR-
maximal routing with transition-aware reranking (GMR-T) cuts transitions by approximately 22% relative to distance-adaptive routing and spectrum assignment
(DA-RSA) at a 3% throughput cost, while bounded-detour TPAR (BD-TPAR) cuts transitions by approximately 11% at only a 1% cost. Deployment pattern also
matters: contiguous HCF rollout lowers transitions by approximately 40% on average while improving carried traffic, reducing the benefit of aggressive
transition-aware routing. These results support BD-TPAR as a practical default under fragmented deployment, GMR-T as a lower-complexity alternative, and TPAR
or GFJ only where the external cost of transitions is high.
arXiv:2607.14543v1 Announce Type: new
Abstract: Vision-language models (VLMs) are increasingly used as the reasoning backbone of embodied agents, enabling robots to interpret visual scenes, follow language instructions, and plan multi-step actions. In household environments, however, safety depends not only on recognizing objects, but also on how actions change the physical scene over time. Existing embodied safety evaluations largely focus on static risk recognition, unsafe instruction refusal, or final-state task completion. As a result, process-level safety failures induced by spatial relations such as support, containment, and proximity remain insufficiently studied. To address this gap, we introduce SAFERELBENCH, a spatial-relation-aware safety benchmark with 507 executable evaluation samples, including 248 spatial-relation samples and 259 non-spatial control samples. Using SAFERELBENCH to evaluate seven open- and closed-source VLM-driven embodied agents, we find a substantial gap between task success and process-level safety compliance: models often complete the requested task while violating process-level safety constraints. Unlike prior benchmarks, SAFERELBENCH explicitly tests whether agents satisfy safety conditions before risk-prone actions, making spatial relations a core dimension in embodied safety assessment. More broadly, our results show that safe embodied intelligence requires not only stronger perception and planning, but also reliable reasoning about how object relations shape risk during interaction.
arXiv:2604.12605v2 Announce Type: replace-cross
Abstract: A time-dependent theory for the interactions between spatially separated lossy cavities in a homogeneous background medium using quantized quasinormal modes (QNMs) is presented. The cavities interact via a bath of traveling photons, described by non-bosonic operators that are orthogonal to the open-cavity QNMs. The retarded (i.e., time-delayed) inter-cavity dynamics are fully described by system-bath correlation functions, in which the emission from one cavity appears as the input field for another. Coupling between quantum emitters (described as two-level systems), placed inside a cavity or embedded in an external medium, and the electromagnetic field (cavity modes and bath photons) is included in the theory, which gives rise to both bath-mediated and QNM-mediated interactions between the emitters.
arXiv:2607.15194v1 Announce Type: new
Abstract: The volume of clear-air turbulence (CAT) in the atmosphere at flight cruising altitudes is increasing rapidly, posing a growing problem for civil aviation and resulting in reduced confidence in aviation safety. There are limited remote detection capabilities for CAT, since clear air produces no measurable radar return. Lidar has been proposed as a viable detection methodology, and several systems have been demonstrated. However, these systems have to date demonstrated limited detection ranges of less than 15 km. In this work, we propose a novel lidar-based CAT detection methodology that uses Rayleigh scattering and relies on a differential detector measurement to quantify beam spread and thereby estimate the eddy dissipation rate (EDR), which is the international aircraft-independent metric for quantifying aviation turbulence strength. Additionally, we present experimental results demonstrating the validity of the optical efficiency model used in the detection simulations. We show that, under modest assumptions, a size, weight, and power (SWAP) constrained system that implements this method can detect moderate CAT at ranges in excess of 30 km, equating to two minutes of flight time for typical commercial aviation cruising speeds, which represents a substantial range improvement over prior approaches. This is an important advance because-for the first time-it potentially allows the cabin to be secured before the turbulence is encountered, reducing the injury risk to passengers and flight attendants.
arXiv:2607.15202v1 Announce Type: new
Abstract: Annotation quality is a major bottleneck in building reliable and explainable artificial intelligence (XAI) systems for mental health research. In depression-related datasets, labels are often assigned without structured evidence, symptom-level justification, or traceable alignment with the criteria of the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, Text Revision (DSM-5-TR), limiting both transparency and downstream model interpretability. We propose a self-evolving, expert-in-the-loop annotation framework for Major Depressive Disorder (MDD) that combines large language model (LLM)-assisted labeling with expert verification. The framework is intended to support the construction of explainable, DSM-5-TR-aligned datasets rather than to perform clinical diagnosis. It operates in three stages: candidate evidence selection from textual records, criterion-level DSM-5-TR analysis, and case-level synthesis that produces label-level diagnostic and severity annotations. A dual-memory architecture, composed of Example Memory and Reflection Memory, is designed to internalize expert feedback and iteratively improve future annotations without retraining. We describe this mechanism and leave its evaluation across multiple feedback cycles to future work. In addition to final labels, the framework exports clinical evidence, reasoning traces, and edit histories, enabling comprehensive auditability. In a pilot study using expert-reviewed samples, the proposed approach improves annotation consistency and explainability while reducing manual revision effort.
arXiv:2606.29059v2 Announce Type: replace
Abstract: World modeling requires forecasting uncertain futures while preserving information useful for
downstream perception. Existing visual world models often struggle to satisfy both goals:
VAE-based stochastic models operate in low-dimensional reconstruction latents, which can
limit perception performance, while deterministic predictors using strong pretrained features
collapse multimodal futures into a single blurry mean. In this work, we propose FlowWM, a
stochastic world model that performs flow matching directly within pretrained feature space
(e.g., DINOv3). This is challenging because pretrained features are substantially
high-dimensional, making standard diffusion recipes suboptimal. To address this, we
investigate the design choices needed for feature-space flow matching and introduce a
differentiable one-step projection mechanism that enables efficient training with temporal
consistency and task-driven objectives. We evaluate FlowWM on two benchmarks: a synthetic
benchmark for systematic evaluation of accuracy and diversity, and a real-world benchmark
FuturePerception. FlowWM improves perception performance, mode coverage, and horizon
robustness, validating our proposed design for stochastic world modeling in high-dimensional
feature spaces.
arXiv:2605.27053v2 Announce Type: replace
Abstract: We revisit the three-body problem in quantum mechanics in two and three dimensions, generating both exact eigenvalues and eigenvectors of the Hamiltonian and a series of approximate solutions as calculated with a variety of different schemes to separate heavy ("nuclear") and light ("electronic") particles. We show that, with minimal extra cost, one can go beyond the Born-Oppenheimer approximation by performing electronic structure calculations parameterized by both the nuclear position (${\mathbf X})$ and the nuclear momentum ($\mathbf{P}$), a so-called phase space theory of electronic structure. In particular, we demonstrate that such phase space electronic structure calculations correctly incorporate the non-inertial Coriolis and centrifugal forces felt by electrons in a moving nuclear frame, thus leading to far more accurate eigenenergies and electronic angular momenta than has been possible before. We also demonstrate that our approach naturally incorporates and generalizes the Moody-Shapere-Wilczek magnetic monopole for the non-abelian Berry curvature (now allowing for vibrational motion rather than a diatomic of fixed length). We argue that the resulting approach should be extremely useful for propagating dynamics where angular momentum flows between nuclei and electrons; in particular, if extended to include spin degrees of freedom, the present approach will offer a practical means to study chiral induced spin selectivity through the lens of chiral phonons and coupled nuclear-electronic motion.
arXiv:2606.31456v2 Announce Type: replace
Abstract: With an increasing number of Object Detection (OD) models being deployed on edge devices, Zero-Shot Quantization for OD (ZSQ-OD) aims to quantize these models when access to the original training data is prohibited. Existing research on Zero-Shot Quantization-Aware Training (QAT) for OD synthesizes training sets through noise optimization. However, this approach struggles to maintain performance in low-bit regions. In this paper, we introduce GoodQ (Generative off-the-shelf models for object detector Quantization), a QAT pipeline that utilizes off-the-shelf generative models to construct a training set. We first identify three challenges that arise when introducing a generative model to the ZSQ-OD task: 1) each image contains dense information with multiple instances, 2) the class-wise distribution in the original dataset is imbalanced, and 3) the pseudo-labels assigned to the generated images can potentially act as noisy signals during QAT. GoodQ addresses these challenges by 1) introducing an Information-Dense Prompting strategy to generate multi-instance images, 2) applying Intrinsic Distribution-Aware Selection to match the pretrained class distribution, and 3) employing Teacher-guided Adaptive Noise Reduction to mitigate noise arising from the QAT process. Our framework achieves state-of-the-art performance in low-bit ZSQ (W4A4) and extends quantization to extreme bit-widths (W3A3). Furthermore, we conduct an extensive analysis to uncover the underlying factors contributing to the efficacy of GoodQ.
arXiv:2607.14360v1 Announce Type: cross
Abstract: Food webs have been extensively studied from both ecological and mathematical aspects. However, most of the models studied in this area do not capture the effects of infectious diseases simultaneously. Recently, the idea of including an infectious disease in a food web model has been investigated. We study and simulate a small food chain consisting of only prey, predators, and apex predators governed by the generalized Lotka-Volterra equations, and we implement the Susceptible-Infected-Recovered (SIR) model on only one of the species at a time in the food chain. To study the effects of an infectious disease on the food chain, we introduce a new parameter that increases the predation rate by a factor of $w$ and decreases the hunting rate by a factor of $1/w$ for infected species. When the infectious disease is present in predators, we observe that predators do not become extinct under any set of parameters; however, an oscillation in their population size occurs under some circumstances, which we do not observe in ordinary SIR or the generalized Lotka-Volterra equations alone. When an infectious disease is present in apex predators, oscillations in the population size do not happen; but if the set of parameters is in a specific range the apex predators may become extinct. Furthermore, the chance of survival of the community, known as community persistence, increases for the predators and decreases for the apex predators.
arXiv:2607.14756v1 Announce Type: new
Abstract: This research investigates the potential of Vision-Language Models (VLMs) to infer building typologies: Construction, Current Use, and Storeys from Google Street View (GSV) images. Predictions generated by VLMs are compared with inference by human experts (civil engineers and architects) as a source of manually labelled ground-truth data. We evaluate several state-of-the-art VLMs, including GPT-4o, Claude 3.5 Sonnet, and Gemini 2.0 Flash. By applying different scaling strategies and prompting techniques, we found that Chain-of-Thought prompts provide an overall more stable model performance. We also investigate the reasoning behind VLMs' building-typology predictions by examining the probabilities of keywords appearing in AI explanations. This enabled us to analyse patterns in these reasonings and identify key themes driving both agreements and disagreements between VLM and expert labels. We find that AI tends to focus on visual indicators, whereas human experts place greater emphasis on broader contextual cues and domain knowledge, in addition to visual cues. Overall, VLM can approximate experts' capability in building-typology classification at scale, with an average accuracy of approximately 70%. The study demonstrates the VLM's potential for AI automation in tasks that require pattern recognition and object identification in an urban context. AI have the potential to serve as complementary and collaborative tools for urban analysis, leveraging their strengths in understanding visual patterns. This study contributes to the exploration of the efficiency and scalability of AI visual prediction and provides insights into the reasoning processes that could support automation processes in urban analysis and prediction.
arXiv:2506.08297v2 Announce Type: replace
Abstract: Attention is the critical component of a transformer. Yet the quadratic computational complexity of vanilla full attention in the input size and the inability of its linear attention variant to focus have been challenges for computer vision tasks. We provide a mathematical definition of generalized attention and formulate both vanilla softmax attention and linear attention within the general framework. We prove that generalized attention disperses, that is, as the number of keys tends to infinity, the query assigns equal weights to all keys. Motivated by the dispersion property and recent development of Mamba form of attention, we design Scalable and Efficient Mamba like Attention (SEMA) which utilizes token localization to avoid dispersion and maintain focusing, complemented by theoretically consistent arithmetic averaging to capture global aspect of attention. We support our approach on Imagenet-1k where classification results show that SEMA is a scalable and effective alternative beyond linear attention, outperforming recent vision Mamba models on increasingly larger scales of images at similar model parameter sizes.
Beyond Implicit Force: Evaluating Explicit Force-Torque Proxies in Action Chunking with Transformers
arXiv:2607.14578v1 Announce Type: new
Abstract: Contact-rich manipulation requires policies to infer interaction state from signals that are often weakly observable through vision and kinematics alone. Action Chunking with Transformers (ACT) has shown strong performance in fine-grained manipulation, but many deployments collect demonstrations through leader-follower teleoperation, where tracking error between commanded leader motion and executed follower motion implicitly encodes contact, resistance, and constraint violation. This paper examines whether ACT's apparent force-awareness depends on this hidden interaction cue. We introduce an observation-centric ACT variant that predicts future follower joint states instead of leader commands, thereby removing the teleoperation-induced discrepancy signal while preserving the rest of the learning pipeline. We then evaluate whether simple joint-torque proxies, derived from onboard motor current or joint effort, can recover contact-aware behavior without external force/torque sensors. Across four real-world tasks spanning surface following, insertion, stiffness discrimination, and force-based stopping, removing the implicit cue leads to severe failures in force-critical phases. In contrast, torque-augmented policies recover robust contact behavior and improve the base ACT policy. These results demonstrate that, on real hardware, the implicit teleoperation cue is a recoverable source of force-awareness, where torque signals are available, a simple proxy matches, surpasses, or further enhances it.
arXiv:2607.14408v1 Announce Type: new
Abstract: A self-evolving agentic loop repeatedly proposes a tweaked version of an agent (its prompt template or program) and accepts or rejects the change based on a per-iteration quality signal. Designing that signal is often the costly part of the project: a reliable scalar reward requires domain expertise and labeled examples that are themselves as expensive to assemble as the agent's underlying task. We propose replacing the scalar at the accept/reject gate with a pairwise validator: a frozen LLM that, given the parent and child candidate, returns a binary verdict on which is better. Pairwise judgment is generally easier and more stable than absolute scoring, due to its contrastive nature, which mitigates the need for strict scale calibration. The validator also requires no training of its own. We integrate the validator into three published self-evolving engines (GEPA, ADRS, ShinkaEvolve) and report two flavors: Adaptive Focus, which retains the engine's existing val-set parent selection, and Soft Elo, which lets the validator's verdicts drive parent selection so that val-set rewards drop as well. Across multiple agents and two artifact substrates (prompt and code), our method matches or exceeds the full-reward baseline on the majority of settings we evaluate, and the pattern survives a cross-family validator swap. The pairwise gate is thus a drop-in replacement for per-step reward design at competitive task accuracy without the labeling cost.
arXiv:2604.20032v2 Announce Type: replace
Abstract: More than half of the Top 500 supercomputers employ GPUs as accelerators. On GPU-accelerated platforms, developers face a key diagnostic gap: profilers show source lines where stalls occur, but not why they occur. Furthermore, the same kernel may have different stalls and underlying causes on different GPUs. This paper presents LEO, a root-cause analyzer for NVIDIA, AMD, and Intel GPUs that performs backward slicing from stalled instructions, considering dependencies arising from registers as well as vendor-specific synchronization mechanisms. LEO attributes GPU stalls to source instructions with the goal of explaining root causes of these inefficiencies. Across 21 workloads on three GPU platforms, LEO-guided optimizations deliver geometric-mean speedups of 1.73$\times$--1.82$\times$. Our case studies show that (1) the same kernel may require different optimizations for different GPU architectures, and (2) LEO's structured diagnostics improve code optimization with large language models relative to code-only and raw-stall-count baselines.
arXiv:2510.24891v2 Announce Type: replace
Abstract: Large language models (LLMs) have demonstrated significant potential to accelerate scientific discovery as valuable tools for analyzing data, generating hypotheses, and supporting innovative approaches in various scientific fields. In this work, we investigate how LLMs can handle the transition from conceptual research ideas to well-structured research plans. Effective research planning not only supports scientists in advancing their research but also represents a crucial capability for the development of autonomous research agents. Despite its importance, the field lacks a systematic understanding of LLMs' research planning capability. To rigorously measure this capability, we introduce the Idea2Plan task and Idea2Plan Bench, a set of benchmarks built from ICML 2025 and Nature Mental Health papers released after major LLM training cutoffs. Each benchmark instance includes a research idea and a grading rubric capturing the key components of valid plans. We further propose Idea2Plan JudgeEval, a complementary benchmark to assess the reliability of LLM-based judges against expert annotations. Experimental results show that GPT-5 achieves the strongest performance on the benchmark, though substantial headroom remains for improvement. Our study provides new insights into LLMs' capability for research planning and lays the groundwork for future progress.
arXiv:2607.10079v3 Announce Type: replace
Abstract: Digital Adoption Platforms (DAPs) are embedded overlays widely used on web systems to guide users through operations inside a page, helping them get started with unfamiliar interfaces quickly. Completing a real task, however, rarely means clicking a few buttons on a single page: it takes a sequence of actions that unfolds across changing page states. Prior studies have also treated automated web agent actions and guide text generation as two separate problems, and most of them feed models textual page representations such as the DOM or accessibility trees rather than the rendered screens that humans actually operate on. In this work we introduce MAG, the first benchmark that unifies task execution and guide writing into a single Multimodal Action and Guide task, with two grounding schemes over screenshots: Set-of-Mark element selection and raw pixel coordinates. We further build a complete harness for this compound task, covering annotation with LLM assistance and human verification, training, evaluation in live environments, and joint metrics for actions and guides. With this harness we evaluate frontier API models and open multimodal models, and report detailed analyses. Finally, we design a GRPO training method augmented with expert trajectories, which nearly doubles the success rate of a supervised 9B agent (from 6.9% to 13.2%) and improves guide quality at the same time. Even the strongest model completes fewer than 40% of the tasks, leaving ample room for future research.
arXiv:2607.14160v1 Announce Type: new
Abstract: Wildfire detection from satellite imagery is a semantic image segmentation problem that has proven to be difficult due to challenges such as class imbalance, feature complexity, and atmospheric interference. In this paper, we build on the foundational U-Net image segmentation model to develop a quantum-hybrid solution in hopes of more effectively modeling the high-dimensional spectral feature space of the Sen2Fire dataset. We inject a variational quantum circuit in the bottleneck portion of U-Net, specifically the QuFeX and QB-Net ansatzes. We test a classical Feature Pyramid Network (FPN) for further comparative analysis of the model, and we also explore classical improvements to the U-Net model and its training process, including a compression of parameters, alternative loss functions, and uniform mixing of input data. Our primary finding is that under matched conditions, both QB-Net (with an $F_1$ score of 31.18) and QuFeX ($F_1 = 30.79$) outperformed the classical U-Net baseline results ($F_1 = 28.71$). Additionally, the classical FPN achieved a comparable score of 31.13. A crucial finding was that data mixing removed a significant domain shift between the geographically-separated train and test sets, which boosted the classical FPN $F_1$ score to 39.76. We validate the architecture's robustness and generalizability to the wildfire detection problem via cross-dataset transfer on the California Burned Areas (CaBuAr) dataset. Overall, we find that quantum machine learning has potential to provide an advantage in the problem of wildfire image segmentation, and further experiments will continue to validate and expand upon this finding.
arXiv:2607.14571v1 Announce Type: new
Abstract: We introduce \emph{gate-zero growth}, a function-preserving (FP) operator for continual learning that adds new residual blocks through a zero-initialised gate. Under a transversality condition, gate-zero growth induces \emph{rank separation} in the functional Jacobian: old directions are unchanged, new-weight directions are exactly flat at the growth point, and new gate directions are the only first-order source of new functional variation. As gates open during continual learning, function drift is $O(\|\boldsymbol{\alpha}\|^2)$ and Jacobian leakage $O(\|\boldsymbol{\alpha}\|_\infty)$, giving a controlled departure from the FP locus. On a $300\mathrm{M}\to857\mathrm{M}$ Transformer adapted from WikiText-103 to BookCorpus, gate-zero growth reaches near-zero old-domain forgetting ($\Delta_A < 0.1$) under both exact-preservation (Isolation) and joint-frontier (Freeze-Nothing) operating points, while a non-FP control ($G_{\text{stack}}$) suffers an order-of-magnitude larger forgetting under the same recipe. The same geometric analysis covers LoRA, ReZero, and zero-init adapter constructions, establishing gate-zero growth as the canonical instance of a shared local geometry that governs safe capacity activation in CL.
arXiv:2607.14658v1 Announce Type: new
Abstract: While Multimodal Large Language Models (MLLMs) excel in general tasks, rigorous scientific reasoning remains challenging due to the limitations of monolithic, linear planning. Such sequential designs often suffer from visual-semantic misalignment, long-context hallucinations, and brittle execution under fixed task granularity. We propose TopoAgent, a self-evolving topological framework that replaces linear trajectories with dynamic, state-isolated graph evolution. TopoAgent first employs a front-end decomposer to fracture complex queries into visually-grounded atoms. These atoms are organized into a Directed Acyclic Graph (DAG) based on their dependencies, enabling strict context isolation to shield the reasoning engine from irrelevant historical noise. Furthermore, we introduce adaptive atomic fission, which dynamically splits bottleneck nodes into finer-grained sub-atoms at runtime when tool capability boundaries are exceeded. Extensive experiments across mathematics, physics, and chemistry benchmarks demonstrate that TopoAgent significantly outperforms state-of-the-art linear agent frameworks, providing a robust, noise-resistant, and self-correcting paradigm for autonomous scientific reasoning.
arXiv:2607.14148v1 Announce Type: new
Abstract: Dance Dance Revolution and In the Groove are rhythm games consisting of songs and accompanying choreography, referred to as charts. Players press arrows on a device referred to as a dance pad in time with steps determined by the song's chart. The process of manual chart generation is timestaking and difficult, motivating interest in automation. We propose ITGPT, a new transformer based architecture for the generation of DDR/ITG charts, and demonstrate significant improvements to generation accuracy and computational cost in comparison to predecessor work.
arXiv:2607.14531v1 Announce Type: new
Abstract: Community structures are common in real networks, and extracting them provides valuable insight in applications ranging from drug discovery to market segmentation. Overlapping community detection (OCD) is the task of clustering networked data in which nodes may belong to multiple clusters. Existing OCD algorithms often struggle to achieve a suitable balance between detection quality and scalability. We, therefore, propose Highway, a scalable OCD algorithm that exploits the sparse backbone of the input network to perform efficient community inference. We used 728 Lancichinetti-Fortunato-Radicchi benchmark networks to compare Highway and its ablated version against 10 existing OCD algorithms. Our results, based on five performance measures, demonstrate a competitive performance for Highway. It ranks first in overlapping normalized mutual information with a 6.9% improvement over the strongest baseline. It also ranks second in all the other four performance measures. These comparative results suggest that Highway coupled with its backbone procedure offers a suitable accuracy-efficiency trade-off. The Highway algorithm is open-source and available as part of the CDlib library.
arXiv:2607.14720v1 Announce Type: new
Abstract: Deep learning models for prostate MRI-based cancer grading may encode clinical covariates that either reflect useful disease-related signal or non-generalising shortcut information, but their role is usually assumed. We propose a causal-reasoning framework for probing covariate dependence in MRI-based International Society of Urological Pathology (ISUP) Grade Group prediction. Rather than treating mpMRI as a direct cause of grade, we model MRI appearance and ISUP grade as observations of latent tumour pathology, and test whether candidate clinical variables act as nuisance correlates, disease-related proxies, or irrelevant covariates in the learned representation. We implement this using an adversarial framework that suppresses the decodability of individual clinical covariate at a time while preserving MRI-based grade prediction. The approach is developed and evaluated on 2,903 prostate MRI examinations, with external validation on 576 patients. We report a set of interesting and previously under-explored imaging-to-clinical-variable interactions in the context of deep learning generalisation. For examples, in binary ISUP Grade Group $\geq2$ classification, suppressing age, BMI, and alcohol use improved AUC by 1.23%, 0.84%, and 1.42%, respectively (all p < 0.05), suggesting reduced non-generalising covariate information; In contrast, suppressing PSA and prostate volume degraded AUC by 1.91% and 7.61% (all p < 0.001), indicating that these variables carried task-relevant signal. These findings show that adversarial covariate suppression can provide a practical representation-level analysis for distinguishing potentially harmful dependence from informative signal in prostate MRI grading models.
arXiv:2604.26979v2 Announce Type: replace
Abstract: In-memory computing (IMC) is a paradigm that enables neural network inference by computing analog matrix-vector multiplications (MVM) directly in memory crossbar arrays, with the potential for energy efficiency gains over conventional von Neumann architectures. In this work we present a simulation framework for N-ary crossbar architectures that retrieves MVM results with minimal implementation assumptions. The XOR and MNIST classification tasks were successfully inferred using a simulated crossbar array of (4x4) 4-states magnetic tunnel junctions (MTJ). MNIST accuracy reached 93.56% (vs. 97.56% software baseline). PCA dimensionality reduction was shown to drastically lower the number of required operations and improve the software baseline, for only a modest reduction in crossbar inference accuracy. We identified weight quantization as the primary error source, and studied its impact alongside systematic non-idealities and random noise. We find that cell-specific random noise is less detrimental than systematic errors due to averaging across the array. Finally, we demonstrate an optimal number of states per cell that balances quantization error against resistance state resolution to minimize total MVM error.
arXiv:2606.11430v2 Announce Type: replace
Abstract: Mathematical knowledge is split between bibliographic databases (e.g., MathSciNet, zbMATH Open) and formal proof libraries (e.g., Lean mathlib), preventing unified access between published results and their formalizations. We propose a relational bridge-database that aligns publication metadata with formal artifacts, providing an interoperability layer between mathematical literature and machine-verifiable proofs. We introduce a paper-level formalization score that measures how much of a publication is covered in formal systems. As a feasibility study, we show how such scores can be estimated via cross-document alignment between informal texts and Lean formalizations, enabling large-scale analysis of formalization coverage. This framework is a first step toward integrating bibliographic and formal mathematical ecosystems into scalable, machine-actionable knowledge graphs linking publications to formal proof objects.