Forskningsradar

Science Journals

Peer-reviewade publikationer — 56237 artiklar

arXiv (fysik & CS) Preprint 2026-07-14
BatteryLake: Agentic, Physics-Grounded Curation of Heterogeneous Battery Aging Data and Benchmarking
arXiv:2607.09762v1 Announce Type: new Abstract: Public battery aging datasets are a critical asset for advanced health management, but their practical use is often limited by inconsistent formats, unclear schemas, and metadata scattered across repositories and publications. Current curation remains largely manual and hard to reproduce, while general-purpose data integration tools miss the domain-specific semantics of electrochemical time-series data. We present BatteryLake, a governed data lakehouse that turns raw public battery data into benchmark-ready assets through an agentic, physics-grounded curation framework, with three contributions. First, LLM agents extract metadata and synthesize dataset-specific converters, grounding every output in verbatim evidence and abstaining when none supports a value. Second, a human-in-the-loop mechanism frames verification as selective prediction and gates admitted data through 26 schema, statistical, and physical-plausibility rules. Third, we release an open benchmark of 41 datasets from over 25 institutions, with standardized SOH and RUL tasks, three split protocols, and eight baseline model families. The platform, benchmark, and curation protocol are publicly available at https://tianwen1209.github.io/batterylake/.
arXiv (fysik & CS) Preprint 2026-07-14
A Conceptual Architecture for Educational Digital Twins Supporting AI Literacy Across Educational and Professional Settings
arXiv:2607.10013v1 Announce Type: new Abstract: In the AI Literacy for Multidisciplinary Professional Readiness and Outreach (AIM-PRO) project, we are creating integrated methods to improve the education on AI literacy. One concept on which the project relies is educational digital twins, that is, digital representations of educator trainers, teachers, and learners that can be used in different stages of the educational process. Such digital twins enable the simulation, monitoring, and optimization of learning experiences. This paper presents the AIM-PRO project and its conceptual foundations, focusing on its core objective: designing and implementing Digital Twins for Education to foster AI literacy across higher education, vocational education and training and professional learning environments.
arXiv (fysik & CS) Preprint 2026-07-14
SyncSpace: Layout-Conditioned 3D Gaussian Splatting for Space Reskinning in Mixed Reality
arXiv:2607.10050v1 Announce Type: new Abstract: We present SyncSpace, a system that achieves both spatial alignment and visual consistency between a generated 3DGS world and physical space. We first scan the space via depth sensing to extract 3D bounding boxes, which we render into a layout-only panorama and feed as a geometric prior to a generative world model, producing a Gaussian splat scene in which objects are re-semantized to fit a target style without per-object control. We then align the generated scene to physical space with a coarse-to-fine registration algorithm, refined manually via pinch gestures when automatic registration does not converge. We demonstrate a hand-tracked engulfment interaction in which the virtual world rises to replace the physical space, and show a single space reskinned into multiple stylistically distinct worlds with its layout preserved.
arXiv (fysik & CS) Preprint 2026-07-14
Functional Expansion Tallies of Matrix Operators for Prediction for Integrated Autocorrelation Time in Batch Monte Carlo: an Analytic 2D Scattering Chain Benchmark
arXiv:2607.10758v1 Announce Type: new Abstract: We investigate functional expansion tallies as a reduced-basis representation for predicting inter-cycle correlations in Monte Carlo transport. Using an analytic two-dimensional isotropic scattering-chain benchmark with reflective boundaries, we compare a conventional discrete-cell Markov-chain estimator with a Galerkin reduced-order model built directly from Monte Carlo tallies of basis-function products. The reduced model estimates integrated autocorrelation time without first constructing a large discrete transition matrix. For the benchmark problem, the cosine basis converges rapidly to the exact result, while polynomial bases show systematic convergence with increasing order. Compared with discrete binning, the reduced-basis approach achieves lower bias at comparable or lower solve cost, suggesting that functional-expansion representations can provide an efficient path toward correlation prediction, uncertainty quantification, and future variance-reduction methods in Monte Carlo criticality calculations.
arXiv (fysik & CS) Preprint 2026-07-14
LLM-Centric Agentic AI for UAV Swarms: Architecture, Enabling Technologies, and Open Problems
arXiv:2607.09756v1 Announce Type: new Abstract: Uncrewed Aerial Vehicle (UAV) swarms have significant potential for applications such as Search and Rescue (SAR) and environmental monitoring, but their real-world deployment is limited by a lack of situational awareness, intermittent connectivity, and significant cybersecurity risks. Agentic Artificial Intelligence (AI) represents a shift from standalone Large Language Model (LLM) toward closed-loop cognitive architectures that integrate perception, memory, reasoning/planning, and action to enable adaptive, goal-directed swarm behavior. Within this framework, Agentic AI provides a unifying structure for autonomous and adaptive swarm operations while expanding the system attack surface compared to conventional AI systems. This paper proposes LLM-Centric Agentic AI for UAV Swarms (LAUS) and reviews key enabling technologies such as onboard and edge computing, 5G/6G connectivity, multimodal intelligence, and cybersecurity mechanisms, and analyzes threats such as Priority Manipulation Attacks (PMA) that can distort decision-making and degrade network performance. Finally, it identifies open research challenges, including hallucination-resistant reasoning, onboard LLM deployment under SWaP constraints, and standardized security benchmarks for perception-reasoning attacks in agentic UAV systems.
arXiv (fysik & CS) Preprint 2026-07-14
Trivial Prompt Reframing Bypasses Safety Guardrails in Google\'s MedGemma-4B
arXiv:2607.09804v1 Announce Type: new Abstract: Open-weight medical language models are increasingly used as the base of patient-facing and clinician-support applications. Their model cards prohibit specific behaviors -- recommending exact drug dosages, issuing definitive diagnoses, prescribing treatments, adjudicating drug-drug interactions, and advising that emergency care can be skipped -- yet a model card describes intended behavior, not robust behavior. We quantify that gap for MedGemma-4B-it under attacks that require no technical sophistication. We build a fully factorial benchmark of 5 guarded-behavior concepts x 50 deterministically templated questions x 6 lay-accessible attack manners x 3 repetitions (4,500 generations), serve the model locally through Ollama under default sampling, and code every response refuse/hedge/comply with three independent judges (an LLM judge, a transparent regex judge, and an NLI-entailment judge). Under the primary LLM judge the overall Attack Success Rate (ASR, the fraction coded comply) is 38.0%. The two framings that reinterpret the request as legitimate dominate: recasting a question as a "medical board exam" item raises ASR from a 29.0% baseline to 53.1% (+24.0 points), and an appeal to an alleged doctor's authority raises it to 43.7% (+14.7); crude instruction-override prefixes have no significant effect. Robustness is dominated by topic: the drug-interaction guardrail is nearly absent (83.2% ASR) while the emergency-deferral guardrail is strong (4.7%) -- and the authority framing is the only attack that breaches it. We report Wilson confidence intervals, cluster-bootstrap effect sizes, a cluster-robust logistic regression, Cochran's Q, per-manner McNemar tests, and inter-judge reliability (Fleiss' kappa = 0.26); absolute ASR is judge-dependent while the ordering of attacks and topics is not. Our findings motivate stronger deployment-time guardrails for open medical models.
arXiv (fysik & CS) Preprint 2026-07-14
JEPA for AI-Native 6G: Predictive Representations and Open Challenges
arXiv:2607.09798v1 Announce Type: new Abstract: Sixth-generation (6G) networks are moving toward AI-native operation, where learning modules are embedded across the radio access network (RAN), edge, and core. This transition requires learning from limited labels, heterogeneous wireless and network data, partial observations, non-stationary propagation, and latency-constrained control loops. Joint-embedding predictive architecture (JEPA) is a promising self-supervised paradigm for this setting because it predicts missing or future representations in latent space instead of reconstructing raw measurements or using contrastive negative samples. This article presents a wireless-oriented tutorial on JEPA for 6G intelligence. We define the JEPA training mechanism, describe how CSI, beam measurements, KPIs, topology graphs, and sensing observations can be tokenized and masked, and position the learned encoder as a predictive representation layer for RAN, O-RAN, edge, and core functions, with task-specific heads or controllers producing final decisions. Then we present an illustrative, beam-management case study suggesting that a wireless-aware target, specifically an auxiliary future beam-energy target during self-supervised pretraining, can improve label efficiency and robustness across shifted deployment conditions relative to a supervised source domain. Finally, we outline open challenges in multi-timescale prediction, action-conditioned modeling, distributed training, trustworthiness, efficient deployment, benchmarking, and standardization.
arXiv (fysik & CS) Preprint 2026-07-14
Reinforcement Learning with Verifiable Physics: Post-training LLMs with Continuous Rewards
arXiv:2607.10474v1 Announce Type: new Abstract: Partial differential equations (PDEs) are foundational to modeling in science and engineering, but constructing reliable numerical solvers remains labor-intensive, demanding expert knowledge of discretization schemes, stability conditions, and boundary treatments. Recent work has begun to frame PDE solving as a code-generation task for large language models (LLMs), yet existing approaches operate primarily at inference time: relying on prompting, debugging, self-refinement, and test-time scaling rather than adapting the model itself. In parallel, reinforcement learning with verifiable rewards has emerged as a post-training paradigm for code and math reasoning, but its verifiers are typically binary: a compiler runs, or a test passes. Such signals discard the graded structure of scientific correctness, where two solvers may both execute and yet differ in solution accuracy by orders of magnitude. In this work, we introduce RLVP: Reinforcement Learning with Verifiable Physics, an RL post-training framework for multi-PDE solver code generation. RLVP addresses this verifiability gap with a hybrid verifier: hard program-validity checks ensure executability, while continuous physics rewards score function-space accuracy and PDE-residual consistency. A single policy is post-trained across diverse PDE families spanning hyperbolic, parabolic, elliptic, and incompressible-flow systems. RLVP improves over both pre-trained and supervised-only baselines on PDE benchmarks, and shows zero-shot improvement transfer to held-out PDEs. We show that a smaller LLM post-trained with RLVP can outperform prompting a frontier model on in-distribution PDE solver generation. The trained policy shows evidence of compositionality in numerical motifs: it recombines stencils, time-stepping schemes, and boundary-handling primitives learned from the PDEs used in training into generated solvers for unseen PDE problems.
arXiv (fysik & CS) Preprint 2026-07-14
IRONSmith: A Visual Dataflow Design Environment for AMD Ryzen AI NPUs
arXiv:2607.10944v1 Announce Type: new Abstract: Machine learning inference increasingly relies on specialized hardware accelerators for throughput and power efficiency. Neural Processing Units (NPUs), such as the AMD Ryzen AI NPU, offer significant ML advantages over CPUs and GPUs, but programming them requires expertise in specialized frameworks. We present IRONSmith, the first visual dataflow design environment for programming AMD Ryzen AI NPUs. IRONSmith provides an interactive canvas displaying the AI Engine tile grid as visually connected blocks, allowing users to design ML dataflow applications by connecting tiles with wires representing FIFOs, split/join patterns, broadcast connections, and DDR transfers without writing any code. Compute kernels are assigned from a pre-built library, and worker functions are configured through property panels. IRONSmith's backend pipeline automatically translates the visual design into executable IRON Python, handling structural completion, import resolution, and dependency management automatically. Generated code executes directly on the AMD Ryzen AI NPU. We demonstrate IRONSmith across ML designs of increasing complexity, from a single-tile vector passthrough to multi-tile matrix operations to a complete Multi-Layer Perceptron, all designed visually and successfully executed on the AMD Ryzen AI NPU. IRONSmith serves educators, students, ML researchers, and engineers by bridging the gap between ML knowledge and NPU programming expertise, widening access to hardware that is rapidly becoming standard across consumer and enterprise devices.
arXiv (fysik & CS) Preprint 2026-07-14
Spectral Origins of the Self-Correction Blind Spot in Autoregressive Generation
arXiv:2607.09803v1 Announce Type: new Abstract: Large autoregressive language models exhibit a self-correction blind spot: they reliably fix identical errors when attributed to an external source yet fail to fix the same errors in their own outputs. Prior work has documented this phenomenon empirically, through controlled error injection, error-depth decompositions, RL-based verifier-corrector training, and intrinsic self-verification, but offers no formal model of why generating a token suppresses the ability to detect its error, no quantitative activation condition for correction markers, and no convergence guarantee for reinforcement-learning-based self-correction. We close these gaps with SPARC, a spectral-algebraic theory of self-correction in autoregressive generation. We define the error-propagation operator as the product of per-step attention Jacobians on the residual stream and prove that the blind spot arises if and only if the spectral radius of this operator is at least one. We derive a sharp activation threshold, given as a function of the spectral radius, that a correction marker must exceed, recovering the 89.3\% blind-spot reduction observed with a simple ``Wait'' marker. We further prove that RL-based verifier-corrector training converges at a rate proportional to the squared coupling strength over the square root of the number of samples if and only if the verifier-corrector coupling matrix has spectral norm below one, and that this criterion is invariant across residual-stream autoregressive modalities, unifying text LLMs and autoregressive image and video generation. Experiments across four backbones and a visual autoregressive probe validate every theorem, with spectral predictions matching measured blind-spot rates within 3.2\% RMSE.
arXiv (fysik & CS) Preprint 2026-07-14
Large Language Model Enhanced Differentiable Trajectory Planning for IoT-Enabled Autonomous Driving
arXiv:2607.10438v1 Announce Type: new Abstract: Autonomous driving planning is a key component of IoT-enabled intelligent transportation systems, requiring vehicles to generate safe, efficient, and executable trajectories in complex urban environments from multi-source contextual information. While imitation learning (IL) has shown promise on large-scale datasets, IL-based planners still suffer from limited coverage of complex long-tail interactions, weak consistency with downstream constrained refinement, and insufficient use of high level scene semantics under real time constraints. To address these issues, this paper proposes a large language model (LLM) enhanced differentiable trajectory planning framework for IoT-enabled autonomous driving. Specifically, we introduce a surrounding agent centric data augmentation strategy to reorganize sur rounding agent trajectories as additional planning supervision, thereby improving the training distribution without collecting additional raw data. We further design a complexity-aware asyn chronous LLM-based semantic enhancement module to extract scene-related high-level semantic features with controlled online overhead. In addition, a differentiable optimization module is incorporated to refine generated trajectories with explicit residual penalties while backpropagating optimization gradients to the upstream planner. Experiments show that the proposed method achieves the best overall scores of 83.63 and 78.29 on the nuPlan closed-loop nonreactive and reactive Hard20 benchmarks, respectively, and CARLA-ROS tests further verify its online deployment and real time closed-loop execution capability.
arXiv (fysik & CS) Preprint 2026-07-14
Deterministic Online Embedding of Metric Spaces into Low Dimensional Spaces
arXiv:2607.10624v1 Announce Type: new Abstract: We study online embeddings of metric spaces into Euclidean spaces of a constant dimension $d>1$, against an adaptive adversary. While the case of $d=1$ is well understood, for higher dimensions little is known. In particular, even for $d=2$ it remains unknown whether the worst-case distortion grows exponentially with the number of exposed points, as it does in the case for the line, or whether it is polynomial, as in the case for unbounded $d$. Our first result is about fixed {\em solid} graphs, i.e., $K_5$, whose edges are solid intervals, equipped with the shortest-path metric. We show that if the input points arrive from such a metric space, they can indeed be online-embedded into ${\mathbb R}^2$ with a polynomial distortion. This refutes the previously believed conjecture that the topological non-embeddability of $K_5$ into the plane could be exploited for establishing exponential lower bounds. The second results is about online embeddings of tree metrics of a certain type, including, e.g., ultrametrics and HST's. Somewhat surprisingly, we show that for metrics from this class the worst-case online embedding into ${\mathbb R}^d$ is not much worse that the offline embedding, both being $n^{\Theta(1/d)}$, and this holds even when $d = \Theta(\log n)$. This is in a stark contrast to the more common situation where the online-offline gap is typically huge, and even exponential. This result allows us to transfer results about probabilistic embeddings of metrics into HST's to low-dimensional Euclidean spaces, in an almost optimal possible manner.
arXiv (fysik & CS) Preprint 2026-07-14
Photonic Theta Cavity: Engineering Bound States in the Continuum in Topological Resonators Beyond the Limitations of Near Field Coupling
arXiv:2607.09915v1 Announce Type: new Abstract: The Theta Cavity is a unique topological resonator architecture which utilizes interferometric coupling to overcome fundamental design limitations associated with near-field evanescent coupling which currently dominates the design space for integrated photonics. The defining device physics is established by the mirror symmetric cross junctions which preserves efficient power transfer between a waveguide and ring resonator creating a strongly correlated phase relationship between the multiple paths. The unique mode selection physics allows for interference driven suppression of radiative pathways enabling strong cavity confinement and the emergence of Bound States in the Continuum (BICs). Analytical models reveal non-Hermitian optical band structure displaying non-trivial topological transitions between BIC and quasi-BIC modes that are robust to attenuation, temperature variations, and typical fabrication non-idealities, which is critical for overcoming intrinsic limitations associated with silicon based integrated photonics. The Theta Cavity architecture also circumvents limitations that arise from proximity requirements of the physical gap used in near-field coupled waveguides which enables a flexible design space for new devices. In this study, we demonstrate phase mediated long-range strong coupling of multiple ring resonators in the Nested Theta Cavity architecture showcasing band structure hybridization resulting in the formation of anti-crossing bandgaps, Dirac crossings, and Fano resonances. The photonic Theta Cavity architecture provides a scalable, topologically robust platform for engineering modes across a multidimensional parameter space with high resilience to perturbation and attenuation, enabling a new approach for the designing of integrated cavity devices.
arXiv (fysik & CS) Preprint 2026-07-14
Actor-Critic Learning for Extended Mean Field Control with Deterministic Policies
arXiv:2607.11005v1 Announce Type: cross Abstract: This paper develops a model-free reinforcement learning framework for continuous--time extended mean field control problems, where both the dynamics and reward may depend on the joint distribution of states and controls. We adopt deterministic feedback policies, under which the state--action distribution is induced directly as a push--forward of the state law. This avoids optimization over stochastic kernels and bypasses key limitations of existing approaches in extended mean field settings. We first establish a model--free sensitivity formula for parameterized McKean--Vlasov dynamics and use it to derive a deterministic policy gradient formula expressed through an advantage--rate function on the Wasserstein space. We then refine this formula by introducing local value and advantage--rate representations that depend on the state, action, and joint state--action distribution, yielding a policy gradient that includes both action derivatives and measure--derivative terms with respect to the control distribution. These characterizations lead to a martingale--based learning principle and motivate a continuous--time deep deterministic policy gradient algorithm combining particle approximations, measure--dependent neural networks, temporal--difference learning, and exploration in either action or parameter space. Numerical experiments on stochastic Cucker--Smale consensus control and optimal liquidation with trade crowding demonstrate the efficiency, stability, and robustness of the proposed method, including problems with explicit dependence on the control distribution.
arXiv (fysik & CS) Preprint 2026-07-14
Knowledge Graphs Meet Graph Neural Networks: A Comprehensive Survey
arXiv:2607.09666v1 Announce Type: new Abstract: Graph Neural Networks (GNNs) have emerged as a powerful paradigm in Knowledge Graphs (KGs) due to their intrinsic ability to model graph-structured data. However, there remains a lack of a systematic review about GNN-based methodologies across the entire knowledge graph technologies pipeline. To address this gap, we first propose a novel two-level taxonomy framework for GNN-based knowledge graph technologies: the KG technologies pipeline and GNN-based perspective. Specifically, the knowledge graph technologies pipeline covers knowledge graph construction, knowledge graph embedding, knowledge reasoning and knowledge graph applications. Meanwhile, the GNN-based perspective provides a new categorization of knowledge graph technologies with GNN models, such as GCN, GAT, and HGNN. Then, we analyze the advantages of GNN technology based on the characteristics of different tasks in the knowledge graph lifecycle. Furthermore, we detailed review various GNN-based models for knowledge graph following the proposed taxonomy, and summarize strengths and limitations. Finally, we discuss unresolved challenges and outline promising directions for future research.
arXiv (fysik & CS) Preprint 2026-07-14
Constraints on unimodular diffusion models with latest observables
arXiv:2607.09750v1 Announce Type: new Abstract: Cosmological models incorporating a time-dependent equation of state have recently been explored \cite{DESI:2025fii}, showing a preference for a dynamical dark energy component. In this work, we investigate a scenario in which an effective, time-dependent cosmological constant arises as an emergent manifestation of a violation of energy-momentum conservation. In \cite{Landau:2022mhm}, such a violation of energy conservation was studied as a diffusion mechanism affecting matter (dark and baryonic), leading to an effective dark energy component within the framework of unimodular gravity. Here, we present an updated analysis using the more recent Type Ia supernova data set from the Dark Energy Survey (DESY5) and the baryon acoustic oscillation (BAO) measurements from the Dark Energy Spectroscopic Instrument (DESI) Data Release 2 (DR2), along with the CMB temperature, polarization, and lensing data from Planck 2018. Our results identify a transition phase that occurs at intermediate times, with slight evidence in favor of the model relative to the $\Lambda$CDM according to the $\mathrm{\Delta DIC}$ criterion. Interestingly, a non-decisive preference for an evolution corresponding to either a time-decreasing or time-increasing effective cosmological constant is found. However, slightly higher values of $H_0$ favor a time-increasing effective cosmological constant. Although the $H_0$ tension is not significantly alleviated, these results suggest that a more refined modeling of the physics of the diffusion mechanism may offer a viable route toward addressing the current discrepancy in the Hubble expansion rate, while also providing a natural framework for incorporating a dynamical dark energy and addressing the problem of vacuum energy contribution.
arXiv (fysik & CS) Preprint 2026-07-14
SciML in the Wild: A Diagnostic Study of When Structural Priors Help and When They Hurt
arXiv:2607.09684v1 Announce Type: new Abstract: Scientific Machine Learning (SciML) methods such as Neural Ordinary Differential Equations (NODEs), Physics-Informed Neural Networks (PINNs), and Universal Differential Equations (UDEs) are most effective when structural priors reflect reliable governing dynamics. We ask what happens when this assumption is violated. Using macroeconomic forecasting as a stress-test domain, we evaluate five model families, ARIMA, LSTM, NODE, PINN, and UDE, across 23 countries using sparse annual data, multiple temporal splits, and five random seeds. Our results show that none of the evaluated models achieve consistently strong forecasting performance, highlighting the difficulty of low-frequency macroeconomic prediction. However, a clear relative hierarchy emerges: less-constrained models, particularly ARIMA and NODE, consistently outperform more-constrained heuristic-prior models such as PINN and UDE. Rather than treating this as a rejection of SciML, we interpret it as a diagnostic result: structural priors can act as misregularizers when they do not match the data-generating process. We identify failure modes including prior misalignment, regime shifts, structural breaks, and optimization instability, and argue that SciML practitioners should test whether structure helps before assuming that more structure is beneficial.
arXiv (fysik & CS) Preprint 2026-07-14
EvoClawBench: Can Agents Learn Reusable Skills from Their Own Runs?
arXiv:2607.09711v1 Announce Type: new Abstract: Existing agent benchmarks primarily test task completion, tool use, or skill utility, but do not isolate whether a runtime can convert evidence from its own runs into reusable skills that improve fresh executions after authoring overhead. We introduce EvoClawBench, a benchmark for this closed-loop skill-learning question on repeated, fixture-backed tasks. EvoClawBench compares direct execution without skills, PreSkill authoring before execution, and PostSkill summarization from first-run evidence followed by a fresh second execution. The suite contains 100 tasks and 502 sub-problems across coding, data, office, security, operations, and domain-document workflows, with support for multiple agent runtimes. Experiments with OpenClaw and nanobot under local execution show that direct baseline performance is strongly runtime-dependent: OpenClaw remains below 20% across models, while nanobot ranges from 56.45% to 96.13%. Self-authored skills have mixed effects. nanobot GPT-5.4 stays above 96% in all modes and MiniMax-M2.7 improves from 90.97% to 94.50% under PostSkill, but nanobot DeepSeek-V4-Pro drops from 77.77% to 4.80% with PreSkill and 0.99% with PostSkill. OpenClaw shows similarly non-monotonic behavior, with some skill runs near baseline and others collapsing. These results indicate that learning reusable skills from an agent's own runs is selective and cost-sensitive, rather than an automatic benefit of adding skill authoring to an agent loop.
arXiv (fysik & CS) Preprint 2026-07-14
Feedback-Coupled Memory Systems in Continuous Time
arXiv:2607.09714v1 Announce Type: new Abstract: The Feedback-Coupled Memory Systems (FCMS) architecture formalizes closed-loop coordination through four abstract operators, two of which - the agent update operator $f_i$ and the environmental update operator $\Psi$ - are left axiomatically undefined in the original framework. To address this, $f_i$ is defined by Mechanism-Based Intelligence (MBI), where agents update locally through a decentralized price mechanism and economic principles, and $\Psi$ is defined by the Coupled Memory Graph Process (CMGP), a non-Markovian framework where the environment is treated as a physical substrate that records and responds to trajectory history coherently without external forcing. The resulting continuous-time FCMS instantiation achieves Lyapunov global dissipativity governed by the computable threshold $4\beta^2 < 2\eta\mu\gamma^2$. This generalizes both the discrete FCMS stability condition $4\eta\beta^2 < \gamma$ and CMGP's physical bifurcation threshold $\alpha_c = 1/K$, confirming that memory dissipation must outpace feedback gain as a universal organizing principle. Numerical simulation with $N=2$ agents and mean-field validation at $N=10^6$ confirm the stability threshold and the self-reinforcing coordination cascade that emerges when it is violated.
arXiv (fysik & CS) Preprint 2026-07-14
A Foundation Model for Multimodal Event Sequences in Financial Applications
arXiv:2607.09955v1 Announce Type: new Abstract: Predictive modeling is a core component of modern financial services, where a wide range of tasks are traditionally addressed using separate models trained on manually engineered tabular features. This task-specific approach limits reuse and makes it difficult to fully exploit heterogeneous data sources such as transaction histories and digital interaction signals. In this paper, we present an approach based on pretraining a foundation transformer model on multimodal sequences of user events. Events from multiple data sources are unified into a single chronological sequence, enabling early fusion of heterogeneous modalities and learning of general-purpose representations via a next-event prediction objective. These representations are combined with existing engineered user features, on top of which lightweight neural models are trained for multiple downstream tasks. The proposed system outperforms traditional task-specific models while reducing development overhead. The approach was deployed in production at one of the biggest banks in Eastern Europe, resulting in measurable improvements in business metrics.
arXiv (fysik & CS) Preprint 2026-07-14
Reconstruction of a dark energy model for the Dirac-Born-Infeld scalar field with the Hubble and DESI data via Gaussian process
arXiv:2607.09731v1 Announce Type: new Abstract: In this study, we reconstruct the dark energy (DE) as a Dirac-Born-Infeld (DBI) scalar field from the Hubble dataset (32 CC + 26 BAO) and the DESI dataset using the Gaussian process (GP). As the GP is a non-parametric and model-independent way to reconstruct a function and its derivative using the data, our reconstruction of the DE equation of state, the DE density parameter, and the potential does not assume any particular model of cosmology. Using Monte Carlo realizations of the GP-reconstructed expansion history, we derive a posterior estimate of the Hubble constant, obtaining $H_0 = 69.53 \pm 2.68$ km s$^{-1}$ Mpc$^{-1}$. This method offers a fully model-independent estimate of $H_0$, relying only on data and GP priors, and provides an unbiased intermediate value useful for reassessing the Planck-SH0ES tension. Using the reconstructed profiles of the scalar potential as a function of the field $\phi$, along with their associated uncertainties, we perform a chi-square curve fitting procedure to assess the viability of four different scalar field potentials, such as Exponential, Power-law, Free Field (quadratic), and Higgs-like potential. This allows us to identify which potential best fits the reconstructed data. We also employ MCMC analysis to place quantitative constraints on the model parameters associated with each potential. Furthermore, we do a $\chi^2$ analysis for all four potentials and comment on the goodness of the fit for each of them. Finally, we discuss possible generalizations of our model-independent framework and outline the phenomenological implications of our findings.
arXiv (fysik & CS) Preprint 2026-07-14
Chalito: An Extensible Library for Filtering-Based State Estimation in Quadruped Robots
arXiv:2607.09968v1 Announce Type: new Abstract: State estimation is essential for quadruped robots, enabling robust locomotion, navigation, and control. While many estimators have been proposed in the literature, existing implementations are often tied to specific robots or software stacks, making fair comparisons difficult. This lack of a general-purpose benchmarking framework hinders reproducibility and slows down algorithmic innovation. In this paper, we introduce Chalito, an extensible MATLAB/Python library for benchmarking filter-based state estimation algorithms in quadruped robots. Chalito imports robot models directly from URDF, supports multiple filtering approaches, and is designed to be easily extended with new methods. The framework runs on both simulated and real datasets, enabling systematic evaluation across robots and filters. To the best of our knowledge, this is the first open-source library exclusively dedicated to benchmarking filtering algorithms for quadruped robots.
arXiv (fysik & CS) Preprint 2026-07-14
ARIMA: Reconstruction-Grounded Predictive Representation Learning for Symbolic Music
arXiv:2607.10003v1 Announce Type: new Abstract: Self-supervised learning for symbolic music has advanced largely through token-level pretraining, but such representations remain tied to tokenizer-specific sequences and often provide time-span-level embeddings only indirectly. In this paper, we propose ARIMA, a reconstruction-grounded latent predictive framework for symbolic music that learns compact window-based representations directly from data. ARIMA encodes each fixed-duration window into a continuous latent representation, trains a causal predictor with contrastive next-latent prediction, and grounds the encoder through structured reconstruction of music elements. This design preserves local musical details while modeling temporal progression across windows. We evaluate ARIMA on downstream tasks spanning various levels of music understanding. Results show that ARIMA is particularly efficient and effective on tasks involving harmonic, timing, and cross-performance retrieval, while remaining competitive with much larger baselines on other tasks. Ablations further show that next-latent prediction is essential for temporally integrated representations, and that structured reconstruction stabilizes latent learning without requiring explicit variance regularization. The code is at https://github.com/AndyWeasley2004/symbolic_music_wm.
arXiv (fysik & CS) Preprint 2026-07-14
Model Predictive Coolant Allocation for Integrated Tab-Surface Cooling of Battery Cells
arXiv:2607.10872v1 Announce Type: new Abstract: Battery electrical tab cooling is effective at reducing internal thermal gradients by exploiting the high thermal conductivity of the current collectors, whereas surface cooling is effective at reducing temperature rise because of its large heat transfer area. Using either strategy alone, however, limits the achievable trade-off between thermal uniformity and temperature rise reduction. This work proposes an integrated tab-surface cooling (ITSC) system in which coolant is dynamically allocated among the lateral surface and tab channels. The allocation is formulated as an optimal control problem in which the battery temperature is regulated towards a desired reference and thermal gradients are minimised. To support this formulation, a first-principles coolant model is developed and coupled with battery and valve-actuation models. The resulting optimal coolant-allocation problem is solved using a computationally efficient real-time iteration model predictive control (RTI-MPC) scheme, with a nonlinear MPC serving as a closed-loop performance benchmark. Evaluation results under realistic driving conditions showed that RTI-MPC reproduces the nonlinear MPC thermal response with absolute errors below 0.0035 degC while reducing the computational cost from several seconds to 19.3 ms, indicating strong potential for real-time implementation. Additionally, evaluation of the proposed ITSC system against conventional cooling configurations demonstrates that ITSC achieves the best overall trade-off between temperature regulation and thermal gradient reduction.
arXiv (fysik & CS) Preprint 2026-07-14
Quantification of Electron Energy-Loss Spectra
arXiv:2607.10693v1 Announce Type: new Abstract: This manuscript summarizes the recent developments in EELS quantification flow as will be implemented in the CEOS Panta Rhei and TEMDM software. This should serve as a technical reference for the algorithms used in the software.