Numerical Analysis
$p$ and $rp$-Spectral Element Methods for Two-dimensional Elliptic Boundary Layer Problems
Sonia, Akhlaq Husain and Arshad Khan
We propose $p$ and $rp$-spectral element methods for
elliptic boundary layer problems on two dimensional rectangular domains. To resolve boundary layers, we use $p$-version with a fixed mesh and an $rp$-version with a boundary layer mesh consisting of thin needle-like elements near the boundary layer and coarse elements away from the layer. Stability estimates are derived using non-conforming spectral element functions. The numerical scheme for both the methods is based on minimising a least-squares functional in appropriate Sobolev norms. We construct robust preconditioners to manage the condition number of the normal equations derived from the least-squares formulation. The method is able to approximate boundary layers at a rate $O\left(\frac{\log W}{W^2}\right)$ for the $p$-version and at the rate $O({\epsilon}\alpha^{2W})$, uniformly in $\epsilon$, for the $rp$-version, where $0<\epsilon\leq1$ is the boundary layer parameter, $\alpha<1$ is a constant and $W$ denotes the degree of the approximating
polynomial. Numerical results are provided for model
elliptic boundary layer problems for a range of boundary layer parameters. Simulation results demonstrate the efficiency and robustness of the method in capturing boundary layers on rectangular domains.
Read the full abstract ▾
Software Engineering
HistoRAG: A Citation-Grounded Question Answering Assistant for Teaching with Scanned Local History and Heritage Archives
Hongzhou Duan and Zikun Guo
Teachers who prepare lessons on local history and cultural heritage work from material that is hard to use. The primary sources are scanned books without a text layer, and the supporting records are administrative catalogs released as spreadsheets. A general chatbot answers such questions fluently but without a verifiable source, which is the property a teacher needs most. This paper presents HistoRAG, a question answering assistant that answers from one regional collection and cites a volume and a page for every fact. HistoRAG transcribes each page with a
vision language model and keeps a line level confidence from the
token probabilities. It builds three stores from the same collection: a hybrid text index, a relational catalog database, and a knowledge graph extracted only from entity dense passages. A lightweight router sends each question to the stores it needs, so that counting questions reach the database and relational questions reach the graph. We build a
benchmark of 516 questions over a collection of 36 scanned volumes and the heritage catalogs of the same region, covering statistical, factual, temporal, and multi-hop questions. HistoRAG answers more questions correctly than passage
retrieval baselines and than a graph based
retrieval system, at a far smaller cost per question. The assistant runs behind a chat interface, so a teacher can check any statement against the page it came from.
Read the full abstract ▾
Computation and Language
Predicting Emerging Topics from Outliers: A Prospective Study of Weak Signals in Embedding Space
Evangelia Zve, Gauvain Bourgne and Jean-Gabriel Ganascia
Some documents that
embedding-based topic models initially classify as noise later become founding members of emerging topics. At publication time, however, they appear as scattered points in
embedding space and are difficult to distinguish from ordinary noise without the benefit of hindsight. We study whether such anticipatory outliers can be predicted prospectively, using only information available when a document first appears. We derive labels from the subsequent trajectories of outlier documents, distinguishing those that anticipate new topics from those that reinforce existing topics or remain isolated, and estimate label confidence through agreement across multiple
embedding models. On two French news corpora, anticipatory outliers prove predictable at publication time. Under cross-validation, $F_1$ rises from about 0.77 over the full eligible population to above 0.90 on high-consensus subsets, and remains at 0.76-0.80 under a strictly chronological evaluation. Predictive performance is driven mainly by
geometric features capturing each outlier's position in
embedding space.
Read the full abstract ▾
Information Retrieval
ScalarLens: Numerical Embeddings with Stable Coordinates and Contextual Responses for CTR Prediction
Heng Yao, Tianying Liu, Yulou Shu et al. (+6 more)
Numerical
embeddings for click-through rate (CTR) prediction are built on a convenient but restrictive premise: a scalar has one representation. This premise conflates where a value lies with what it means for the current sample. On the Criteo validation split, the same numerical interval carries residual click evidence with opposite signs across categorical and numerical contexts, even after additive main effects are removed. Production pipelines compound this mismatch because externally normalized features require transformations and statistics to remain synchronized between training and serving. We introduce ScalarLens, a numerical
embedding that preserves what a value is while adapting how it should be interpreted. A monotone local mesh constructs a stable coordinate from the focal scalar alone; bounded low-rank dynamics then produce a contextual response without moving that coordinate or replacing categorical
tokens and the CTR backbone. In a 1,539-run primary evaluation covering 19 representations, three
datasets, nine backbones, and three seeds, ScalarLens ranks first in 25 of 27 settings on original numerical scales and second in the remaining two. Matched ablations show that scale correction, additional local capacity, and generic conditioning do not reproduce the gain. A controlled study further recovers categorical, numerical, and mixed response mechanisms under context shift while the focal coordinate remains exactly invariant. A complete rerun under shared standardization retains significant advantages over DEER, DAES, and NaryDis, showing that the result is not explained by tolerance to raw scales alone. ScalarLens therefore recasts numerical
embedding as a measurement problem: coordinates belong to values, while predictive responses belong to values in context.
Read the full abstract ▾
All authors Heng Yao, Tianying Liu, Yulou Shu, Yong He, Chuan Yuan, Kaibin Qiu, Guowei Chen, Jiayu Zhao, Siyun Hou
Artificial Intelligence
Right Choice of Classification Algorithms Based on Reinforcement Learning for Prediction of Non-Alcoholic Fatty Liver
Hasan Samadbin and Arman Daliri
There are many complex issues in the world of artificial intelligence. Some of these problems are solved using other artificial intelligence methods, which are called artificial intelligence for artificial intelligence. Finding an appropriate classifier
algorithm is a time-consuming task. For this reason, an
algorithm that can automatically learn the choice of classification
algorithms is very important. Classification
algorithms are useful in predicting various diseases. Also, Primary Biliary Cirrhosis is one of the most well-known diseases that have been predicted by classification
algorithms. This research's most significant achievement and novelty is the automatic increase in learning through a scoring method of reinforcement learning is called square learning (SL). In this research, an
algorithm is presented that learns to automatically select the appropriate classification
algorithm to predict Primary Biliary Cirrhosis. In this article, with inspiration from four evaluation metrics in classification
algorithms, a new reinforcement learning method by the name of Fourth Degree Learning has been presented. In this research, we increased the performance of the classification
algorithms used in this method from 63% of accuracy and achieved 98% accuracy.
Read the full abstract ▾
Information Retrieval
X-Rec Technical Report
Chenglei Shen, Chenzhe Huang, Dong Jiang et al. (+15 more)
Recent advances in generative modeling have reshaped recommender systems by formulating recommendation as a next-item generation problem. Existing
retrieval approaches primarily follow two paradigms: user-to-item (U2I) methods represent user context using one or a few deterministic
embeddings, which limits the ability to capture diverse and multi-mode interests, while semantic-ID-based autoregressive (SID-AR) methods model more expressive distributions but suffer from quantization errors and the low throughput of sequential decoding. To address these limitations, we propose X-Rec to directly learn the recommendation distribution in the continuous item
embedding space through flow matching and generate
embedding triggers for approximate nearest neighbor
retrieval. X-Rec incorporates three key designs to make this formulation effective and efficient. First, we introduce anchor conditioning to decompose generation into coarse semantic-region selection and fine-grained refinement. Second, we adopt Riemannian flow matching to align generative trajectories with the hyperspherical geometry of item
embeddings. Third, we design a late-interaction
diffusion Transformer that restricts repeated velocity-field estimation to the final
Transformer layer. On a streaming
benchmark, X-Rec substantially outperforms U2I baselines, matches the
retrieval quality of SID-AR methods, and delivers 3.46x higher inference throughput than SID-AR. X-Rec has also been deployed as a new
retrieval source for a specific vertical content on TikTok, where two consecutive launches have yielded significant improvements in both vertical engagement (+4.1484%) and general engagement (+0.0111%).
Read the full abstract ▾
All authors Chenglei Shen, Chenzhe Huang, Dong Jiang, Hongjie Gao, Jue Zhang, Kun Xú, Lincan Cai, Nan Zhuang, Pan Zhang, Shi Chen, Shunchi Zhang, Xiaoyu Ye, Yang Jin, Yu Zhang, Zhenwei An, Zhongtao Jiang, Zhiwei Wang, Kun Xǔ
Sound
Towards Deployable Underwater Vessel Classification
Abishek Soti, Thura Pyae Sone, Naqib Ibnul et al. (+4 more)
We propose a compact
underwater acoustic classification framework combining multi-representation feature engineering, temporal statistical pooling, and compact convolutional architectures designed for acoustic time-frequency and cochlear representations. We investigate multiple conventional and auditory-inspired representations and first evaluate lightweight classifiers and Conventional Neural Networks (CNNs) on ShipsEar
dataset. On the provided split, a two-layer CNN achieves a macro F1 of 0.9918, while a Radial Basis Function Support Vector Machine (RBF-SVM) reaches 0.9883. However, source-recording provenance cannot be reconstructed, preventing verification of recording-independent generalisation. We therefore evaluate on DeepShip
dataset using recording-level partitioning before
segmentation. Under this protocol, a 157K-parameter compact CNN achieves a test macro F1 of 0.7226, while an 11.17M-parameter ResNet18 provides no improvement in validation performance under the matched setting. These results demonstrate the importance of representation-aware feature and model design, together with rigorous recording-level evaluation, for classification performance and deployability in compact
underwater acoustic systems.
Read the full abstract ▾
All authors Abishek Soti, Thura Pyae Sone, Naqib Ibnul, Htoo Htet Aung, Henry Zhong, Gregory Cohen, Ying Xu
Cryptography and Security
Poster: FedWM-Guard: Thwarting Imagination Poisoning in Federated World Model-based Autonomous Driving
Sheng Liu and Panos Papadimitratos
Federated learning (FL) can improve world model (WM)-based autonomous driving (AD) without centralizing raw private vehicle data, but it also turns model aggregation into a safety-critical integrity boundary. We introduce a new threat in
federated WM-AD, namely imagination poisoning: compromised vehicles submit bounded WM updates that preserve benign short-horizon predictions yet corrupt long-horizon rollouts (e.g., trigger-conditioned) during training, thereby misleading a downstream planner. We present FedWM-Guard, to the best of our knowledge, the first defense to characterize planner-facing rollouts in
federated WM-AD, screen authenticated updates in hidden-canary scenarios, audit predicted futures against later observations, and invoke a WM-independent safety shield when persistent inconsistency is detected. Unlike parameter-space defenses, it scores what an update makes the model imagine, not only how the update looks. We also outline how we plan to evaluate it under non-IID (not independent and identically distributed) data, adaptive attacks, and benign distribution shift. This work highlights an unexplored domain,
federated WM-AD, and its threat surface and potential countermeasures.
Read the full abstract ▾
Rings and Algebras
Duality for matrix space questions
Yuval Wigderson
We present a new proof of a classical
theorem of Dieudonné: if a linear space of $n\times n$
matrices consists entirely of singular
matrices, then its dimension is at most $n^2-n$. Our proof is based on a surprising “duality” argument: we prove this universal upper bound by exhibiting a single
matrix space that serves as a lower bound for a related problem. Interestingly, this approach only works for certain fields, but we use model-theoretic arguments to obtain the same result for all fields. We hope that this approach can be generalized to provide new duality-based proofs of other classical
theorems on
matrix spaces, and give some preliminary results in this direction.
Read the full abstract ▾
Robotics
Anthropomimetic Soft Robotic Forearm with Independently Articulated Carpal Bones Enabling Human-Like Adaptive Stiffness Modulability
Yoshinobu Obata, Yinlai Jiang, Hiroshi Yokoi et al. (+1 more)
The human wrist exhibits adaptive stiffness modulability: joint stiffness anisotropy can be actively regulated through muscle co-contraction. This functionality is essential for stable manipulation, yet the underlying morphological factors remain unclear. To identify these factors, we developed an anatomically accurate anthropomimetic soft robotic forearm comprising eight independently movable carpal bones interconnected by ligaments, 22 actuated muscles, and compliant fingertips. We measured wrist joint stiffness under four muscle activation patterns across three skeletal configurations: anatomically normal carpal bones, a fused proximal carpal row, and a
geometric ellipsoidal skeleton. The stiffness ellipse exhibited low stiffness along the dart-throwing motion (DTM) direction when finger muscles were activated, but high stiffness along the same direction when wrist and finger muscles were activated simultaneously. These results agree with previously reported human measurements, demonstrating that precise anatomical replication reproduces human-like stiffness modulability. Fusing the proximal carpal row eliminated the low DTM-direction stiffness under finger muscle activation, while the
geometric ellipsoidal skeleton showed poor stiffness ellipse reorientation across all conditions. Carpal bone motion analysis revealed significantly opposing coupling patterns between wrist and finger muscles at the proximal carpal row, accompanied by a consistent but non-significant trend at the midcarpal joint, providing a mechanical explanation for this modulation. These findings demonstrate that carpal bone morphology plays a dominant role in human wrist stiffness modulation and provide design principles for
humanoid robot wrists.
Read the full abstract ▾
All authors Yoshinobu Obata, Yinlai Jiang, Hiroshi Yokoi, Shunta Togo
Human-Computer Interaction
A Human-Like Pedestrian Model for Automated Driving Simulations
Ruofeng Wang, Patrick Ebel, Philipp Wintersberger et al. (+1 more)
Automated vehicles must be able to interact with pedestrians safely and efficiently across diverse traffic situations. Although driving simulators offer a scalable testbed for learning such capabilities, existing theory-inspired pedestrian models are narrow in scope and limited to go/no-go crossing decisions in single-lane settings. While data-driven approaches can predict pedestrian behavior in complex situations, they lack sufficient observations in rare, safety-critical scenarios. Here, we propose an approach to training pedestrian models in simulators so that learned policies generate demonstrably human-like behavior in realistic, complex traffic scenarios, including multiple lanes, heavy traffic, and dangerous driving styles. Our technical contribution is a novel definition of pedestrian-vehicle interaction as a partially observable Markov decision process (POMDP) with theory-grounded perceptual, cognitive, and motor constraints. It accounts for the highly adaptive nature of human behavior in traffic and simulates how people adjust their responses according to perceived danger, time pressure, and the complexity of the situation. When trained via deep reinforcement learning (RL) with domain randomization in a simulator, the model reproduces the broadest range of empirical findings shown so far on human crossing behavior, including gap acceptance, yielding acceptance, hesitation, and evasive speed adjustment. We show that learned policies transfer to unseen traffic environments, and can be further adapted to local traffic norms with finetuning. Together, these results establish a blueprint for simulator-ready pedestrian models that can support the development and evaluation of automated driving systems.
Read the full abstract ▾
All authors Ruofeng Wang, Patrick Ebel, Philipp Wintersberger, Antti Oulasvirta
Solar and Stellar Astrophysics
Metal-Poor Stars as Metal-Rich Impostors via Planet Engulfment
Zongheng Li, Shi Jia, Jie Su et al. (+2 more)
Planet engulfment leads to planetary mass
accretion and angular momentum deposition into the
stellar envelope. This study uses MESA models to investigate the long-term impact of engulfment on
stellar metallicity and rotational velocity, as well as the effect of rotational mixing on metal enrichment. Our models show that engulfment occurring during the late main-sequence (MS) stage yields significantly higher surface enrichment than at the ZAMS. Rotational mixing generally suppresses enrichment; however, metal-poor
stars can sustain super-
solar abundances for several Gyr, whereas high-
metallicity stars remain largely unaffected. Although engulfment causes an instantaneous
spin-up, this angular momentum injection is transient, and the
star quickly re-converges to its standard rotational
evolution. Thus, MS
spin-up cannot serve as a long-term observational indicator. Because convective envelope thickness and MS lifetime dictate the magnitude and duration of enrichment, identifying engulfment signatures requires balancing the long-lived but diluted signals in 0.7 M_sun
stars against the strong yet short-lived enrichment in 1.2 M_sun
stars. Consequently, 1.0 M_sun
stars offering both substantial enrichment and a relatively long MS lifetime represent optimal targets for observational searches.
Read the full abstract ▾
All authors Zongheng Li, Shi Jia, Jie Su, Shang-Fei Liu, Jianheng Guo
Human-Computer Interaction
Through Human Eyes and Machine Eyes: Understanding View Mismatch in Video See-Through Extended Reality
Yanming Xiu
Video see-through extended reality (VST XR) systems commonly use headset screenshots or captured frames as proxies for the user's first-person visual context. However, the system-captured view and the user's effective visible field do not necessarily coincide: a screenshot records a rectangular machine-readable frame, whereas the user's effective visible region can be more constrained and non-rectangular. This paper studies this human-system view mismatch in VST XR. We formalize the relationship between the system-captured region and the human-visible region by defining their co-visible, system-only, and human-only regions. \rev{We then conduct a pilot-level boundary measurement on Meta Quest 3, revealing a clear mismatch between the rectangular screenshot frame and the approximate human-visible boundary. Building on this model, we analyze how view mismatch can affect screenshot-based XR sensing and downstream
vision-
language model tasks. Through four representative case studies, we illustrate potential risks and failure modes including
prompt injection,
privacy leakage, human-invisible information bias, and missing human-visible information. Our results show that view mismatch is not only a
geometric artifact, but can also introduce security,
privacy, and reliability concerns for AI-integrated VST XR systems.
Read the full abstract ▾
Software Engineering
Orchestrating AI-Assisted Code Remediation: Socio-Technical Bottlenecks in a Large Industrial Repository
Andreas Bexell, Lo Gullstrand Heander and Emma söderberg
Background: Code degradation in large, long-lived codebases is costly to remediate through manual refactoring and opportunistic clean-ups. LLM-based coding assistants can perform mechanical remediation at scale, but their impact on industrial workflows is underexplored. Objective: We investigate how massive AI-assisted code remediation affects build-on-commit continuous integration (CI), code review, and team coordination in a large industrial repository, and which socio-technical bottlenecks constrain such remediation when source editing becomes cheap through AI assistance. Method: We report on a 15-day exploratory single-case field study in which an experienced developer used a command-line AI coding buddy to remediate widespread issues in a closed-source industrial C++ repository. We triangulate Gerrit metadata with a developer diary and team chat, analyzed through descriptive statistics and qualitative coding. Results: AI-assisted remediation rapidly generated hundreds of commits touching thousands of lines, saturating CI and reviewer
attention. Naïve per-file commits overloaded build-on-commit CI; Switching to directory-based batching and capping the number of files per change restored throughput, but still required explicit review solicitation, negotiation of acceptable commit granularity, and iterative follow-up to resolve build and static-analysis failures. Conclusion: When mechanical editing is cheap, CI capacity, review effort, and change orchestration become primary bottlenecks. Sustainable AI-assisted remediation in very large repositories requires deliberate control of commit, review, and CI batch granularity and treating semantic change sets, such as “fix all instances of warning X”, as first-class units of work that can be sliced differently for developers, reviewers, and CI.
Read the full abstract ▾
Robotics
Representation World Model: Learning States, Transition and Executable Plans in Representation
Yijun Yuan, Weicheng Zheng, Weibang Wang et al. (+7 more)
We propose the Representation World Model (RWM), which learns states, transitions, and executable plans directly in representation space. Unlike existing world models that typically learn latent representations together with explicit dynamics models and perform planning through search, optimization, or policy-based prediction, RWM directly incorporates planning into the learned representation geometry. RWM learns the representation geometry by applying inverse-dynamics supervision locally along latent paths constructed from endpoint representations, requiring these paths to preserve task-relevant state and transition information. At inference, planning is performed by directly constructing a latent path between the current and goal representations, with inverse dynamics used to recover the corresponding actions, without recursive rollouts or action-space search. Experiments on continuous-control
benchmarks demonstrate the effectiveness of RWM for direct planning, while results on robotic manipulation further show its potential to extend to more complex
embodied control tasks. These results suggest that planning directly in representation space provides a promising alternative to conventional world-model planning.
Read the full abstract ▾
All authors Yijun Yuan, Weicheng Zheng, Weibang Wang, Minghui Qin, Chang Sun, Junhao Huang, Kenan Li, Anmin Liu, Yicheng Yao, Hang Zhao
Analysis of PDEs
Sharp stability for the second-order weighted Heisenberg Uncertainty Principle with explicit stability constants and optimizers
Xiao-Ping Chen
By using spherical harmonicas decomposition and Gaussian-type Poincaré inequalities, we establish several sharp stability estimates for the following second-order weighted Heisenberg Uncertainty Principle \begin{equation*} \int_{\mathbb{R}^N} \!\frac{|\Delta u|^2} {|x|^{2\alpha}} \mathrm{d}x \int_{\mathbb{R}^N} |x|^{2\alpha+2}|\nabla u|^2\mathrm{d}x \geq \frac{(N+4\alpha+2)^2}{4} \left(\int_{\mathbb{R}^N} |\nabla u|^2 \mathrm{d}x\right)^2, \end{equation*} and \begin{equation*} \int_{\mathbb{R}^{N}} \frac{|\Delta u|^{2}} {|x|^{2\alpha}} \mathrm{d}x +\int_{\mathbb{R}^{N}} \left|x\right|^{2\alpha+2} |\nabla u|^{2}\mathrm{d}x \ge\left(N+4\alpha+2\right) \int_{\mathbb{R}^{N}} |\nabla u|^{2}\mathrm{d}x. \end{equation*} We also provide the explicit value and the necessary and sufficient condition for attainability of the sharp stability constants. Moreover, when $\alpha=0$, our results reduce into those of [Calc. Var. Partial Differential Equations 64 (2025), Paper No. 129], [J. Funct. Anal. 290 (2026), Paper No. 111321] and [arXiv:2510.00453].
Read the full abstract ▾
Sound
Spot, Separate, and Enhance: Fully Generative Approach for Audio Mixing
Ilpo Viertola, Giulio Cengarle, Gouthaman KV et al. (+2 more)
We introduce Spot, Separate, and Enhance (SSE), the first
multimodal, user-guided generative model for audio remixing and enhancement. SSE enhances
video content by rebalancing the audio, removing unwanted audio sources, and reducing reverberation, guided by both
video and textual descriptions. To support its training and evaluation, we propose DegradedMix, a new
dataset built on the audio remixing
benchmark MuddyMix. We also adopt evaluation metrics from generative modeling, which better capture the creative nature of remixing than standard reconstruction-based metrics. SSE outperforms existing baselines in both controllability and remixing quality, as shown by extensive experiments. Project page: https://sse-ai.notion.site
Read the full abstract ▾
All authors Ilpo Viertola, Giulio Cengarle, Gouthaman KV, Daniel Arteaga, Lie Lu
Superconductivity
Hidden magnetic order within the pressure induced superconducting dome of UTe2
Kaixin Ye, Lubin Wang, Dengpeng Yuan et al. (+12 more)
Unconventional superconductivity typically occurs near
magnetic instabilities, and the corresponding
spin fluctuations are widely believed to play a crucial role in mediating electron pairing. UTe$_2$ is a promising candidate for exhibiting multiple
spin-triplet superconducting phases when tuning with applied pressure and
magnetic fields, but the nature of the magnetism driving these unconventional pairing states is undetermined. Our measurements of UTe$_2$ under applied pressures and
magnetic fields reveal the presence of a
magnetic order hidden within the pressure-induced superconducting dome, which vanishes together with the superconductivity once there is sufficiently high pressure to induce the three-dimensional
antiferromagnetic phase. Extrapolation of the phase boundary of the hidden
magnetic order, which is most likely
antiferromagnetic in nature, points to a zero-temperature
quantum critical point that coincides with the maximum transition temperature of the pressure-induced superconducting dome, suggesting that it could corresponds to the parent
magnetic phase of the critical
antiferromagnetic spin fluctuations driving the triplet superconductivity. These findings advance the understanding of the interplay of magnetism and superconductivity in an exemplar candidate triplet
superconductor, which is necessary for revealing the microscopic origin of the different unconventional superconducting phases.
Read the full abstract ▾
All authors Kaixin Ye, Lubin Wang, Dengpeng Yuan, Binbin Zhang, Yanan Zhang, Ye Chen, Yu Liu, Xin Lu, Chaofan Zhang, Qiuyun Chen, Shiyong Tan, Frank Steglich, Lin Jiao, Michael Smidman, Huiqiu Yuan
Artificial Intelligence
IndicBankBench: Evaluating Safety and Reliability of Language Model Assistants in Indian Retail Banking
Suvradip Paul, Chandra Bhushan, Harsh Sharma et al. (+4 more)
Banking assistants must use account-specific information to answer requests and, in many cases, take actions through tools. Evaluating only the final response misses important errors. An assistant may ask for information it already has, rely on stale context, select the wrong account, or write an invalid value after stating the correct one. We introduce IndicBankBench, a 799-case
benchmark for Indian retail banking spanning five operational domains, a capability/refusal domain, and twenty primary axes. Cases are evaluated at four stages: safety, action and tool use, response adequacy, and advisory quality. Tool use and most safety checks are deterministic. A narrow resolver handles only ambiguous confirmation-before-write cases, while a separate LLM judge evaluates semantic response adequacy. We run every case three times and report strict pass^3, which requires success on all trials. Across the eleven evaluated models, strict reliability ranges from 43.7% to 58.2%, whereas at-least-once success ranges from 60% to 74%. This gap shows that at-least-once success can overstate dependable banking behavior. The case-level diagnostics also distinguish systems that ask unnecessary questions from those that act but fail to reconcile customer context or fully resolve the request. We release the cases, mock environment, and evaluation harness.
Read the full abstract ▾
All authors Suvradip Paul, Chandra Bhushan, Harsh Sharma, Nitin Kukreja, Yatharth Dedhia, Keyur Doshi, Prashant Devadiga
Robotics
HarnessPAI: An Evolving Harness for Physical AI
Xin Wang, Wenhao Wu, Menghao Zhang et al. (+20 more)
Physical AI aims to build
embodied agents that perceive the world, understand and reason about it, and decide how to act. Yet the field has focused primarily on the last component: the action model that maps observations to low-level controls. The prevailing training recipe can erode the perceptual and
reasoning capabilities needed for robust behavior, leaving even strong action models vulnerable to scene perturbations and long-horizon tasks. We introduce HarnessPAI, a model- and embodiment-agnostic Harness framework for Physical AI that treats code as the executable and evolvable interface that organizes the underlying action primitive. The framework separates two timescales: within a rollout, it executes open-loop at the program level, with a fixed program guiding and checking execution; across rollouts, it evolves closed-loop, using execution feedback to revise the program and distill failures into reusable skills. Across desktop
robot arms, household
robots, a
robot vacuum, and a legged walking
agent, HarnessPAI demonstrates consistent improvements against baselines. Beyond the harness, the converged program also acts as a low-cost, reliable source of expert data: it generates training data for fine-tuning VLA and WAM models on tasks where it is hard for these models to obtain successful trajectories on their own. Our results suggest that the frontier of Physical AI depends not only on stronger action models, but also on executable harnesses that integrate perception, task understanding and
reasoning, and action execution into a unified, verifiable, and feedback-driven system. Website: https://darwin-
agent.github.io/HarnessPAI
Read the full abstract ▾
All authors Xin Wang, Wenhao Wu, Menghao Zhang, Zhi Wang, Kun Shao, Jian Luan, Yang Li, Qing Li, Shangding Gu, Huichi Zhou, Shuqing Shi, Fei Ni, Shuo Lu, Weicheng Meng, Kang Li, Jin Wu, Kang Zhao, Shangmin Guo, Gen Li, Yongqiang Tang, Zhizhong Zhang, Yuan Xie, Heng Qu
Astrophysics of Galaxies
Optical spectropolarimetry of extreme H$\alpha$ line profiles in seven active galactic nuclei
J. Biedermann, F. Marin, D. Hutsemékers et al. (+1 more)
Some active
galactic nuclei (AGNs) are known to show extremely asymmetric broad Balmer lines, with their main peak redshifted or blueshifted by thousands of km s$^{-1}$, severely challenging our understanding of the standard AGN paradigm. We wanted to explore the causes of such asymmetric features by carrying out a spectropolarimetric study of a sample of bright AGNs with known extreme Balmer line profiles. We present optical spectropolarimetry of seven bright (V $<$ 17.5) Seyfert-1
galaxies obtained with the Very Large
Telescope (VLT) FOcal Reducer/low dispersion Spectrograph 2 (FORS2) instrument from March 2012 to January 2013. The main broad Balmer lines (H$\alpha$ and H$\beta$, and H$\gamma$ occasionally) were sampled with a resolving power ranging from 350 to 500. We performed a detailed spectral decomposition of each H$\alpha$ profile to isolate kinematic substructures within the broad line region (BLR) and measured the polarization degree and angle of each component, providing constraints on the inner geometry and scattering properties. The broad H$\alpha$ components exhibit various scattering geometries. Their polarization as a function of full width at half maximum (FWHM) reveals two distinct behaviors, corresponding to either a BLR of roughly constant scale height or a stratified structure with decreasing height at higher velocities. This supports a disk-wind geometry linking equatorial and polar regions.
Read the full abstract ▾
All authors J. Biedermann, F. Marin, D. Hutsemékers, C.M. Gaskell
Signal Processing
Enhancing 5G NTN VSAT RACH in GNSS-Denied environments
Francesco Menzione, Alejandro Gonzalez-Garrido, Flavien Ronteix-Jacquet et al. (+1 more)
The 3GPP 5G non-terrestrial networks (NTN) technology fundamentally relies on a global
navigation satellite system (GNSS) fix at the user equipment (UE) to pre-compensate for user-link delay and Doppler shifts prior to any transmission, including the random-access procedure for initial access. In GNSS-denied or interfered environments, this dependency triggers preamble collisions and connection dropouts, undermining satellite-based global connectivity. To address this vulnerability, this paper introduces a novel single-satellite positioning framework designed for higher-frequency (Ku- or Ka-band) 5G NTN very small aperture terminals (VSATs) equipped with a phased-array
antenna. The multi-step scheme derives an initial coarse position from angle of arrival (AoA) observables, extracts frequency of arrival (FoA) and time of arrival (ToA) measurements from low Earth orbit (LEO) downlink reference signals, and hybridizes them within a Levenberg-Marquardt
navigation filter. Systematic evaluations across static grids and realistic dynamic urban trajectories detail the incremental gains of transitioning from single-epoch AoA error minimization to multi-epoch angular tracking, followed by the integration of Doppler and ranging observables. Grounded in a realistic error budget, the proposed framework reduces initial positioning uncertainty from tens of kilometres down to a few kilometres. This level of positioning accuracy, combined with the GNSS-resilient features introduced in 3GPP Release 20 NR NTN specifications, enables robust initial access in compromised operational environments with minimal relaxation of 3GPP standards.
Read the full abstract ▾
All authors Francesco Menzione, Alejandro Gonzalez-Garrido, Flavien Ronteix-Jacquet, Ottavio M. Picchi
Machine Learning
Edge AI on Constrained Devices for Binary Sleep-Wake Classification in Dynamic Environments
Stefan Reitmann and Lena Oden
This paper presents an Edge AI-based system for detecting sleep and wake states in non-stationary mobile environments using resource-constrained embedded hardware. Conventional approaches relying on accelerometer-based activity metrics are highly susceptible to motion and vibration artifacts and are limited by strict compute and energy budgets of wearable and IoT devices. To address these challenges, a
multimodal pipeline is designed and implemented on an ESP32-S3 microcontroller. The system combines inertial sensing for head movement analysis and visual pose classification. A dual-core architecture with FreeRTOS enables parallel execution of real-time data acquisition and on-device inference. Sleep detection follows a two-stage strategy: low-movement detection over a temporal window, followed by visual validation of poses. Experimental results show accuracies of 96.5% for motion-based detection and 89% for pose classification, yielding robust binary sleep-wake classification. Field tests confirmed feasibility in representative mobile scenarios. The results demonstrate that
privacy-preserving, local sleep detection is achievable on edge hardware through careful co-design, while highlighting limitations in sensing intrusiveness,
dataset scale, and system integration.
Read the full abstract ▾
Astrophysics of Galaxies
Binary-pulsar Timing: A 3D Solar-System Accelerometer and Unknown-Source Probe
Zheng-Long Wang, Zi-Qing Xia, Bo Zhang et al. (+1 more)
The acceleration of the
Solar System barycenter (SSB) offers a unique precision probe of the local gravitational environment, enabling searches for otherwise invisible gravitating sources such as
Planet Nine and nearby (
primordial) black holes. Using the binary-
pulsar timing measurements, we develop an analysis framework that combines the line-of-sight differential accelerations of 26 binary
pulsars to jointly reconstruct the three-dimensional acceleration of the SSB and place directional upper limits. Across three smooth
Galactic-potential baselines, the residual acceleration is consistent with zero. We obtain directional 95% upper limits of $0.198$, $0.262$, and $0.693~\mu\mathrm{as}\,\mathrm{yr}^{-1}$ over 50%, 75%, and all sampled directions, respectively. These limits are tighter by factors of 3.5, 3.6, and 1.6 than a matched
Gaia EDR3
benchmark and improve previous
pulsar constraints by more than an order of magnitude. We further extend the framework to the full SSB–
pulsar two-endpoint response, enabling direct position-dependent mass constraints on unknown gravitational sources. A $10\,M_\odot$ object is excluded within 0.39, 0.34, and 0.21 pc over the same sky fractions. Our unknown-mass limits surpass the
tidal-equivalent INPOP19a reference beyond 0.2 pc by about an order of magnitude over most of the sky at 1 pc.
Read the full abstract ▾
All authors Zheng-Long Wang, Zi-Qing Xia, Bo Zhang, Yi-Zhong Fan
Digital Libraries
A framework for linking literature-based knowledge integration and infrastructure-supported knowledge integration: Opportunities and challenges from a case study
Mahlet Degefu Awoke, Hadi Ghaemi, Lauren Synder et al. (+2 more)
Integrating knowledge across disciplines is central to sustainability research, yet most evidence-synthesis methods rely on findings as reported in publications, limiting verification and reuse of underlying data and workflows. We develop a conceptual framework linking literature-based and infrastructure-supported knowledge integration, using a systematic review case study to examine when integration can extend beyond reported findings. We reviewed 37 studies on
climate change, violent conflict, and household food security. Literature-based synthesis enabled integration across all included studies, whereas access to reusable outputs was limited: over half provided no data availability statement, 27% reported availability upon request, but reusable data and workflows were available for only 8%. To explore infrastructure-supported integration, we used the TIB Knowledge Loom to represent studies with accessible data and code as machine-readable outputs, and produced a knowledge gap map (KGM) from manually extracted and Loom-derived data, comparing manual and infrastructure-supported synthesis. Where outputs were reusable, synthesis could be produced directly from data and workflows rather than from publications. These findings show that literature-based synthesis can be complemented by infrastructure-supported integration where outputs are accessible and usable, and that advancing knowledge integration depends not only on infrastructures but on making data, code, and workflows accessible, executable, and reusable.
Read the full abstract ▾
All authors Mahlet Degefu Awoke, Hadi Ghaemi, Lauren Synder, Markus Stocker, Tilman Brück
Distributed Computing
Cross-Model Autoscaling for Shared LLM Serving
Xin Zhang, Xianyan Xie, Zhen He et al. (+6 more)
Multi-model LLM serving is moving toward shared MaaS clusters, where co-hosted models compete for a fixed GPU budget while each model experiences time-varying demand and must satisfy its own latency SLO. Existing LLM autoscalers remain largely model-local: their signals expose local runtime activity or delayed latency outcomes, and their scale-up or provisioning decisions do not directly determine how shared capacity should be allocated across competing models. We present the
Token-service-share Rebalancing Engine (TRE), a control-plane framework for hot-switched multi-model LLM serving. TRE introduces
Token Service Share (TSS), a calibrated, demand-normalized signal that estimates effective
token service per active or queued request and yields a comparable health score across heterogeneous models and SLO classes. Guided by TSS, TRE coordinates bounded receiver–donor capacity movement under a fixed GPU budget: it separates fast rescue from slower rebalancing and incrementally reallocates active replicas toward models with the largest calibrated service deficits. We implement TRE on a Kubernetes-based hot-switch serving stack without modifying the inference scheduler. Across seven LLM serving traces, TRE reduces P95 end-to-end latency by 11.9–79.0% and P99 latency by 12.5–72.6% compared with a state-of-the-art KV-cache-based reactive autoscaler running on the same hot-switch runtime. The gains hold on both targeted stress probes and production-derived conversation/code traces, where TRE reduces P95/P99 latency by 50.8/63.7% and 79.0/72.6%, respectively. These results show that effective hot-switched autoscaling requires not only fast replica actuation, but also calibrated service-deficit signals and coordinated cross-model capacity arbitration. Our code and artifacts are available at https://github.com/zxzx9898/
Token-service-share_Rebalancing_Engine.
Read the full abstract ▾
All authors Xin Zhang, Xianyan Xie, Zhen He, Xijin Yin, Xingtong Lin, Bangbo Liang, Zequn Cheng, Peihao Huang, Guo Chen
Dynamical Systems
Functional dynamic mode decomposition: Learning infinite-dimensional systems from data
Stefan Klus and Eirini Ioannou
Dynamic mode decomposition (DMD) is a data-driven method that computes the best linear approximation of the underlying dynamical system and decomposes the dynamics into a superposition of characteristic spatiotemporal patterns. Originally introduced by the fluid dynamics community, DMD and its extensions have found widespread use in many other research areas such as molecular dynamics,
climate science, engineering, finance, and neuroscience. Applications include dimensionality reduction,
forecasting, system identification, control, and spectral clustering. In order to apply DMD to partial differential equations, the spatial domain is typically first discretized using finite difference or finite element techniques, thus implicitly rendering the problem finite-dimensional. We extend projected and exact DMD to infinite-dimensional systems. Rather than estimating
matrices from vector-valued observations, our DMD variants learn finite-rank operators from functional data such as observables, densities, or wavefunctions. We show that conventional DMD
algorithms can be regarded as special cases of their functional DMD counterparts. All results will be illustrated with the aid of guiding examples. We focus in particular on Koopman, Perron-Frobenius, and Koopman-von Neumann operators associated with graphons, ordinary differential equations, and
stochastic differential equations.
Read the full abstract ▾
Analysis of PDEs
A Morrey-to-Lebesgue Equivalence for Convolution Calderón–Zygmund Operators
Rishad Shahmurov and Veli Shahmurov
We prove that, for vector-valued convolution Calderón–Zygmund operators, boundedness on a single nontrivial Morrey space is equivalent to the corresponding global $L^p$ boundedness. Thus one Morrey scale already contains the full finite-$p$ boundedness information. The implication from Morrey to $L^p$ is obtained by a separated-copy amplification argument that reconstructs the global norm from a single scale-local estimate; the converse is proved in the same vector-valued framework by a local/far-field decomposition. As a consequence, boundedness of the vector-valued
Hilbert transform on one nontrivial Morrey space is equivalent to the UMD property. The result shows that Morrey estimates do not bypass the classical
Banach-space obstruction: they detect it exactly.
Read the full abstract ▾
Robotics
OREN-X: Octree Residual Network for Real-Time Multi-Modal Mapping
Zhirui Dai, Qihao Qian, Dinh Minh Nguyen et al. (+6 more)
To achieve general-purpose autonomy over long horizons, a
robot needs to maintain spatial environment information that supports a variety of tasks: geometry for planning and control, radiance for rendering and relocalization, and
vision-
language features for open-vocabulary grounding. Existing methods represent and estimate each modality separately, multiplying
memory and compute cost while forgoing potential synergy among the representations. We develop OREN-X, an online mapping method that uses an octree in 3D space as a shared data structure for indexing and storing a multi-modal field, capturing
geometric, radiance, and
vision-
language information. OREN-X provides efficient unified storage and
retrieval of these data in explicit/implicit and full/compressed form. Our unified representation yields cross-modality synergy: SDF estimates are sharpened by occupancy and radiance, while GPU-based ray-octree traversal and octree query enable real-time rendering. We also use online dictionary learning to compress the
vision-
language features, shrinking them 3.7x below full per-vertex storage while raising the query accuracy. On Replica, OREN-X maps in real time (80+ fps for SDF and 30+ fps for all four modalities), improves near-surface SDF accuracy by 33% over single-modality baselines, and improves mean open-vocabulary 3D mIoU by 71% and mean accuracy by 61% over the best prior method.
Read the full abstract ▾
All authors Zhirui Dai, Qihao Qian, Dinh Minh Nguyen, Quan-Dung Pham, Kiana Bronder, Carlos Nieto-Granda, Yiyu Chen, Quan Nguyen, Nikolay Atanasov
Computer Vision
Med-AR: Autoregressive Vision-Language Pretraining for Long-Tailed Chest X-Ray Classification and Uncertainty-Aware Evaluation
Janhavi Prabhu, Sahil, Akshay V et al. (+3 more)
Long-tailed chest X-ray classification requires visual representations that capture both common abnormalities and subtle, infrequent findings. We propose Med-AR-8B and Med-AR-2B, two radiology-native autoregressive
vision-
language models pretrained with structured reports, abnormality-focused text, and region annotations. We evaluate the transfer of their visual encoders to multi-label classification against contrastive, self-supervised, and supervised pretrained encoders, including Med-CLIP, CheXFound, EVA-Base, ARK, and BioViL-T, using a common ML-Decoder classification head. To assess fine-grained recognition, we also construct LLM-expanded, report-derived label sets for MIMIC-CXR and CheXpert. Across PadChest, MIMIC-CXR, and CheXpert, Med-AR-8B outperforms Med-CLIP in mean AUROC and AUPRC for head, medium, and tail findings. On MIMIC-CXR, it increases tail-label mean AUPRC from 0.1033 to 0.1441. Med-AR-2B achieves the strongest discrimination results on PadChest. Across the broader encoder comparison, a Med-AR variant achieves the highest mean AUROC and AUPRC in every reported prevalence group on each public
dataset. Both Med-AR variants also achieve lower excess area under the risk-coverage curve than Med-CLIP on all three public
datasets, indicating improved selective-prediction performance under the evaluated protocol. Internal results are metric-dependent, with Med-CLIP retaining advantages in overall and tail AUPRC and in selective prediction. These findings establish Med-AR as a strong pretraining recipe for long-tailed chest X-ray classification on the evaluated public
benchmarks and demonstrate the value of assessing discrimination and selective prediction together.
Read the full abstract ▾
All authors Janhavi Prabhu, Sahil, Akshay V, Shivam Shukla, Manoj Tadepalli, Preetham Putha