Peer Revenue Publish or profit

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

All of arXiv

Friday 4 September

Computation and Language

Reinforcement Learning for improving Large Language Models' Catalan text simplification capabilities

Although automatic text simplification (ATS) is critical for accessibility, its progress has not matched the rapid evolution of broader natural language processing techniques. This paper investigates the application of reinforcement learning (RL) to improve the quality of ATS for low-resource languages using Large Language Models (LLMs). The paper introduces a novel reward function, designed to guide LLMs toward a targeted simplification style with Group Relative Policy Optimization (GRPO), that combines the SARI metric with specific penalty components. The effectiveness of GRPO with this reward function is motivated and demonstrated by post-training IberianLLM-7B-Instruct on the ASSET dataset. After post-training on the English ASSET, the model's ATS performance improves on two curated Catalan benchmarks while also successfully suppressing previously observed negative behaviors. Cross-lingual transfer learning is explored by translating ASSET into Catalan and Spanish and post-training the model on each version, but these fail to show a significant improvement on the out-of-domain benchmark.
Read the full abstract

Machine Learning (Statistics)

Minimax Rate for Diffusion-Guided Local Intrinsic Dimension Estimation based on Fokker-Planck Equation

Diffusion-guided methods have recently emerged as effective tools for probing the intrinsic geometry of high-dimensional data. Understanding their accuracy requires quantifying both (i) finite-scale approximation error relative to the true local intrinsic dimension (LID) and (ii) estimation error from finite samples. We address these two questions for the FLIPD field (Kamkari et al., 2024; arXiv:2406.03537), defined at the population level through the logarithmic scale derivative of a Gaussian-smoothed density and admitting a Fokker–Planck representation. For $\alpha$-smooth densities on a regular $d$-dimensional manifold, we establish a uniform $O(\sigma^2)$ approximation bound relative to the manifold dimension $d$. Under expected integrated squared loss, we prove a minimax lower bound of order $(n\sigma^d)^{-1}$ for estimating the FLIPD field from $n$ observations when $n^{-1/(2\alpha+d)}\le\sigma\le\sigma_0$. A plug-in FLIPD estimator based on the empirical denoising score matching (DSM) solution attains a matching upper bound. Thus the minimax rate is of order $(n\sigma^d)^{-1}$ and it becomes $n^{-2\alpha/(2\alpha+d)}$ at its smallest scale.
Read the full abstract

Superconductivity

On the Stability of the Polar Phase of Superfluid 3He in Nematic Aerogels

We study the stability of the solution corresponding to the polar phase of superfluid 3He in nematic aerogels with respect to perturbations induced by magnetic impurity scattering. We show that when the perturbation simultaneously breaks axial and time-reversal symmetry, the polar-distorted A phase is realized with the highest transition temperature. Within the proposed model of a p-wave scattering potential, which accounts for the effective spin-orbit interaction between the orbital angular momentum of a scattering quasiparticle and the magnetic moment texture of the scatterer, we solve the eigenvalue and eigenvector problem for the Gor'kov self-consistency equation. Finally, we discuss how our results relate to the available experimental data.
Read the full abstract

Cryptography and Security

Cost-Aware Hierarchical Multi-Agent Ransomware Detection and Family Attribution

Sandbox execution and memory forensics are among the most resources intensive stages of malware triage. Static analysis can scale to millions of files while dynamic and memory analysis require minutes per sample. Because multimodal ransomware detectors often apply every modality to every sample, costs and latency rise even when static evidence suffice.We proposea cost aware Hierarchical Multi Agent System (HMAS) that frames evidence acquisition as a budgeted sequential decision problem enabling timely ransomware detection. Specialist agents generate schema validated risk signals for each modality while domain controllers aggregate evidence and coordinate analysis. The orchestrator starts with static evidence and escalates only if confidence is low or signals disagree. A locally hosted language model provides a limited, accoutable review without replacing the deterministic pipeline. Experiments on 16 ransomware families show 94.12% binary accuracy (F1 0.9376, ROC-AUC 0.9989). For multi class family attribution, it achieves 98.90% accuracy (macro-F1 0.9748). The system resolves 57.95% of the samples using static analysis alone whhile only 6.21% require the full pipeline. Compared with exhaustive cost of 12.0 units,the proposed system reduces te average cost to 6.724 units, representing approximately 44% saving. Zero day experiments further examine generalization to unseen ransomware families. Additional evaluations examine cost sensitivity and component level contribution. LLM-assisted review provides only marginal improvement in F1 while introducing additional computational overhead. Overall, the results show the potential of adaptive hierarchical multi agent orchestration to balance ransomware detection performance, family attribution analysis cost and decision latency.
Read the full abstract

All authors Mubashar Iqbal, Hifsa Asif, Saddam Hussain Khan, Umme Zahoora, Asifullah Khan

Computation and Language

A Systematic Comparison of Multilingual Interpretability Methods Reveals Anisotropy-Driven Failures

Multilingual language models develop shared cross-lingual representations, and various interpretability methods claim to quantify this sharing. These methods have been developed largely in isolation, and when they disagree, it is unclear whether the disagreement reflects a property of the model or an artifact of the measurement. We compare four sharing metrics (CKA, ANC, GMM dominance per token, and ILO) across 21 base models from five families (125M-14B parameters) and correlate each with cross-lingual transfer on five downstream tasks. We find that the metrics differ in their quantification of cross-lingual sharing in these models and suggest that the disagreement traces to anisotropy, the tendency of representations to cluster in a narrow cone of the embedding space. Only ILO's correlation with cross-lingual transfer (Spearman's $\rho = 0.90$) survives controls for model size, family, and per-task variation. We therefore recommend ILO as the primary sharing metric, to be reported alongside anisotropy diagnostics.
Read the full abstract

Optics

Modulation-Frequency Dependence of Spatial Resolution in Optical Correlation-Domain Reflectometry

The spatial resolution of optical correlation-domain reflectometry (OCDR) has conventionally been described by an expression that is independent of the modulation frequency $f_m$, source linewidth $\delta\nu$, and receiver resolution bandwidth (RBW). However, our previous measurements showed that the spatial resolution $\Delta z$ improves with increasing $f_m$. Here, we develop a theoretical model for OCDR with a frequency shifter by evaluating the electrical power detected by an electrical spectrum analyzer and explicitly including $\delta\nu$ and the RBW $B$. The model predicts two regimes. At low $f_m$, $\Delta z$ decreases approximately in proportion to $1/f_m$, with the proportionality determined by the combined source and receiver spectral response. At high $f_m$, $\Delta z$ approaches a constant value determined by the modulation amplitude $\Delta f$ and independent of the source linewidth and RBW filter. Measurements at RBW = 1 MHz reproduced the transition between these regimes over correlation orders up to 2048. At RBW = 10 MHz, a Voigt representation of the spectral response overestimated $\Delta z$, whereas direct use of the measured unmodulated beat spectrum reduced the discrepancy to approximately 10 to 20%. These results provide a quantitative description of the modulation-frequency dependence of OCDR spatial resolution and clarify its trade-off with measurement range.
Read the full abstract

Robotics

Strict Modes Everywhere - Bringing Order Into Dynamics of Mechanical Systems by a Potential Compatible With the Geodesic Flow

Strict nonlinear normal modes provide very regular families of oscillations within conservative mechanical systems. However, a strict normal mode will generally be an isolated curve within the configuration space of the system. In this letter, we design a potential that will densely fill the configuration space with strict normal modes such that each configuration belongs to one mode and each mode passes through a common point, the equilibrium. As the potential can be realized by (nonlinear) elastic elements it can be used to execute a variety of periodic trajectories very efficiently. Most of the required torques will come from the elastic elements in the system and not from the actuators. We also design a controller stabilizing the system to a desired target mode and a controller performing swing-up and compensating dissipated energy. Finally, we showcase the approach for a two DoF manipulator. The experiments show that the approach performed well for the example system.
Read the full abstract

Differential Geometry

Deformations of Kähler and Balanced Hyperbolicity

We study the deformation stability of Kähler and balanced hyperbolicity. We prove that Kähler hyperbolicity is open under arbitrary small holomorphic deformations, but need not be closed, even in families of smooth projective surfaces. By contrast, balanced hyperbolicity is not open in general: in every complex dimension at least five, we construct a balanced hyperbolic central fibre with non-balanced nearby fibres. We establish criteria for the deformation stability of balanced hyperbolicity using Aeppli cohomology and Lefschetz theory. For principal holomorphic torus bundles over products of curves of genus at least two, we compute the relevant hyperbolic cohomology, obtain a numerical characterization of balanced hyperbolicity, and prove its openness under small deformations.
Read the full abstract

Machine Learning

Federated Attack Campaign Detection via Contrastive Encoding of Threat Indicators in Gradient Updates

Detecting orchestrated cyberattack campaigns that span multiple organizations traditionally requires sharing sensitive telemetry and threat intelligence across institutional boundaries and country borders, a barrier that Federated Learning removes by training shared threat detectors directly on local data. We propose FedIoC, a modular framework in which clients fold locally available structured threat indicators into their gradient updates; we instantiate the client-side encoder with a supervised contrastive loss over IoC-matched flows. Within each training batch, flows that match any known indicator pattern form the positive set; the contrastive objective pulls their learned embeddings together and pushes non-IoC embeddings away, so that campaign-relevant structure is, by design, expressed in the gradient direction. Clients sharing indicators for the same attack campaign then produce aligned gradient components, which the server clusters by the cosine similarity of their updates to recover global campaign patterns without any direct IoC transmission. We evaluate FedIoC on two public threat-detection benchmarks distributed across FL clients that each observe only a fragment of every active campaign and hold disjoint indicator sets derived from their local telemetry. In this regime the FL server recovers cross-organizational campaign cohorts directly from gradient geometry. We contribute FedIoC as a modular framework for this setting, and use it to pinpoint the non-IID gradient structure as the main driver of recovery and to define the open problem of designing encoders that improve on it.
Read the full abstract

Quantum Gases

Spin-Charge Subordination in the Infinite-$U$ $SU(N)$ Hubbard Chain

We obtain the exact full counting statistics of charge and flavor transport in the one-dimensional infinite-$U$ $SU(N)$ Hubbard model. The no-passing constraint freezes the ordered flavor sequence while the particle coordinates evolve as free spinless fermions. Consequently, charge transfer is governed by a free-fermion determinant, whereas flavor transfer is exactly subordinated to the number of particles crossing the observation cut. For an arbitrary traceless Cartan generator this yields an exact relation between the charge and flavor cumulants. The complete asymptotic flavor distribution is a universal non-Gaussian M-Wright law. Together, these results establish spin–charge subordination as a kinematic mechanism for anomalously slow internal-state transport in one-dimensional impenetrable quantum gases.
Read the full abstract

All authors Cătălin Paşcu Moca, Ovidiu I. Pâţu, Gergely Zaránd, Balázs Dóra

Information Retrieval

Inventory-Grounded Policy-Level Optimization for Training-Free AI Search

Early in deployment, an AI search system typically operates over a frequently updated product catalog, so the available items and their properties cannot be treated as stable knowledge that can be encoded in fixed prompts or strategies. Fine-tuning, reinforcement learning, and static prompt patches fit poorly: labels are scarce, rewards drift with inventory, model releases are costly, and prompt fixes quickly stale. We present Inventory-Grounded Policy-Level Optimization (IGPO), a training-free approach for fixed AI search pipelines. IGPO separates policy from environment facts: it learns Policy Guidelines for acting on runtime inventory evidence rather than memorizing available items. Online, IGPO grounds each query by probing the inventory and constructing an inventory portrait, then injects relevant Policy Guidelines into the retrieval and selection prompts. Offline, stochastic rollouts are grouped by query – mixed outcome groups directly yield contrastive signal, and an inventory-guided exploration loop distinguishes missed retrieval routes from cases where no matching support is found under the observed inventory evidence. Since May 2026, IGPO has been deployed in a commercial smart-assistant AI search system. A 14-day online A/B test of the complete IGPO treatment shows a 3.17% relative CTR lift and a 38.9% reduction in audited bad cases.
Read the full abstract

All authors Wei Zhou, Tiandeng Wu, Jiandong Ding, Zhufeng Fan, Yi Cao

Quantum Physics

Transition between weak and strong measurements in the presence of post-selection

In the weak measurement regime, post-selection can result in the observation of anomalous weak values, seemingly contradicting the eigenvalue statistics observed when the measurement interaction is strong. Here, we investigate the dependence of meter statistics on measurement strength in a post-selected measurement. We find that the meter statistics in the intermediate regime between weak and strong measurements is nearly independent of measurement strength and show that, in this regime, the system performs a measurement of momentum on the meter. Our results show that post-selection is extremely sensitive to the action of the meter on the system, highlighting the difference between the weak measurement regime where this back-action is negligible and the average meter position is a valid measurement result, and the intermediate regime where the meter statistics should not be misinterpreted as a measurement outcome. The transition between weak values and eigenvalues is facilitated by a reversal of the roles of the system and the meter, where the post-selection acts as a readout of information about the meter. The intermediate measurement regime thus reveals that the apparent contradictions between anomalous weak values and eigenvalues originate from the non-classical effects of back-action on the post-selected meter statistics.
Read the full abstract

Materials Science

Challenges in orbital current-driven domain wall motion in light metal/ferrimagnet heterostructures

Recent advances in spintronics suggest that orbital Hall currents generated by charge injection in light metals can provide nonequilibrium angular momentum without relying on strong spin-orbit coupling (SOC). Here, we examine whether such orbital currents can drive domain wall (DW) motion in an amorphous ferrimagnetic alloy, Gd$_{25}$(Fe$_{9}$Co$_{1}$)$_{75}$ (GFC), where the rare-earth sublattice offers strong SOC for orbital-to-spin conversion. We compare three representative heterostructures: Pt/GFC as a spin Hall reference, light-metal (Mn or Ti)/GFC for direct orbital-current injection, and light-metal/Pt/GFC incorporating an ultrathin Pt layer for orbital-to-spin conversion. Whereas Pt/GFC exhibits robust and reproducible spin-orbit-torque-driven DW motion, no current-driven DW motion is detected in Mn/GFC or Ti/GFC. Second-harmonic Hall measurements nevertheless reveal finite damping-like torques in both Mn/GFC and Ti/GFC, demonstrating that angular-momentum transfer into GFC does occur but is far weaker than in Pt/GFC. Inserting a 1-nm-thick Pt conversion layer strongly enhances the damping-like torque and restores DW motion. Thickness-dependent analysis further shows that DW mobility and depinning thresholds correlate with the interfacial Dzyaloshinskii-Moriya interaction and domain-wall width, highlighting weak torque conversion and insufficient interfacial stabilization of chiral DWs as key challenges for orbital-driven DW motion in light-metal/ferrimagnet heterostructures.
Read the full abstract

All authors Min-Gu Kang, Jaerin Kim, Benjamin J. Jacot, Laura Van Schie, Pietro Gambardella

General Relativity and Quantum Cosmology

Dirac fields in LRS III spacetimes: a dynamical systems analysis

Within the $1+1+2$ covariant formalism, we investigate Locally Rotationally Symmetric (LRS) class III spacetimes sourced by a self-gravitating Dirac field expressed in polar form. By recasting the covariant field equations into an autonomous dynamical system, we perform a phase-space analysis of their solutions. We show that the cosmological evolution asymptotically converges to a contracting Bianchi I spacetime, a result also supported by numerical integrations of the system.
Read the full abstract

Artificial Intelligence

CPR-IE:A Compression-Prediction-Resource Intelligence Efficiency Metric

Comparing intelligent systems under deployment constraints requires more than predictiveaccuracy.This paper develops Compression-Prediction-Resource Intelligence Efficiency (CPR-IE) as a protocol-relative ordering by representational economy, predictive quality, and resourceburden. The analysis separates two questions-how raw resource consumption is represented, andhow the resulting attributes are aggregated. Proportional-increment composition uniquely yieldslogarithmic cumulative burden, and context-independent ratio response yields power responsesto compression, prediction, and burden; with reference normalization the representation is I(C,P,T).We prove Pareto consistency, unit invariance, boundary behavior, trade-off identities, ranking-stability regions, and cross-task aggregation. A translog parent model makes interaction restrictions explicit, and further results establish cardinal and ordinal identification, sub-Gaussianfinite-sample ranking guarantees, robust selection under exponent uncertainty, and deterministicregret bounds. Minimum description length, algorithmic complexity, proper scoring rules, varia-tional inference, and Landauer's principle motivate measurement choices but do not entail theformula. CPR-IE is a constructed efficiency representation, not a universal law or a definition ofintelligence itself.
Read the full abstract

Computation and Language

Recurrence Is Not Enough: Causally Validating Multilingual SAE Translation Features in Gemma 2 and 3

Sparse autoencoder (SAE) features are increasingly used to explain and steer language-model behavior, but it remains unclear whether a feature found in one language context plays the same causal role when processing prompts in another language. We study this question using translation-initiation features (Wu et al., 2026). We reproduce the SAE feature discovery method from Wu et al. in Gemma 2 and extend it to multilingual settings that vary prompt language, source language, and target language. We then test whether features that recur across settings affect translation behavior by amplifying or ablating their activations during inference. We also examine whether the method can be applied to Gemma 3. In both models, we observe an identical finding: although we can find more than 20 features that activate frequently across all discovery settings, causal validation shows that nearly all have small or inconsistent effects. In contrast, one feature – Gemma 2's (L10, 5717) and Gemma 3's (L20, 2456) – consistently improves COMET scores when amplified and degrades them when ablated across 23 language settings. These results show that feature recurrence can overstate cross-lingual transfer, while identifying a language-agnostic translation-initiation direction in Gemma 2 and Gemma 3.
Read the full abstract

All authors Giang Son Nguyen, Nhi Ngoc-Yen Nguyen, Wray Buntine, Dung D. Le

Robotics

CoLMIN: LLM-based Multi-Decision Path Negotiation for Cooperative Autonomous Driving

Multi-vehicle cooperative autonomous driving enhances the safety and reliability of autonomous driving systems through information sharing among connected vehicles, demonstrating significant potential for improving traffic safety. LLM-based approaches leverage strong reasoning capabilities of LLMs to enable effective inter-vehicle negotiation and improve cooperative driving performance. However, driving decisions in complex traffic scenarios are inherently multi-solution in nature. As a result, existing negotiation-based methods often converge prematurely to suboptimal solutions, hindering consensus formation and limiting the practical deployment of cooperative autonomous driving systems. To address this challenge, we propose CoLMIN, the LLM-based multi-decision path negotiation framework for cooperative autonomous driving, achieving stable decision consensus through multi-decision path negotiation and reflective reasoning. To achieve stable and high-quality consensus in cooperative autonomous driving, CoLMIN consists of three key components: (i) an LLM-based Multi-Intent Negotiation module (LMin), which adopts a Negotiator-Evaluator paradigm and generates multiple candidate driving intentions for joint evaluation; (ii) an Evaluation-based Shallow Reflection Module (ESRM), which analyzes negotiation outcomes and provides feedback to guide subsequent negotiations, thereby accelerating consensus formation; and (iii) an LLM-based Deep Reflection Module (LDRM), which performs long-term reflection over negotiation histories to mitigate cognitive fixation and prevent the system from converging to suboptimal solutions. Experimental results in the CARLA simulation environment demonstrate that CoLMIN significantly outperforms existing methods in challenging interactive driving scenarios.
Read the full abstract

All authors Zhe Huang, Zhaoxin Fan, Shuo Wang, Wenjun Wu, Xuan Zhao, Min Liu

Artificial Intelligence

When Financial Fine-tuning Fails: A Three-Level Detectability Analysis of Numerical Hallucination in Domain-Adapted Language Models

Financial large language models are increasingly deployed for summarization of reports and disclosures, where numerical hallucination poses significant practical risks. While prior work often attributes such hallucination to insufficient numerical reasoning, this assumption has not been systematically tested under controlled fine-tuning settings. In this paper, we conduct a cost-effective, controlled study of numerical hallucination in financial summarization across three model variants: a base instruction-tuned model, a domain language-adapted model (FT-A), and a numeracy-enhanced domain model (FT-A+B+C). We introduce a three-level detectability taxonomy distinguishing between overt hallucination (currency-denominated fabrication), covert-explicit hallucination (professional-convention numbers), and covert-implicit hallucination (ungrounded quantitative claims). Our results reveal that domain fine-tuning substantially degrades numerical restraint at all detectability levels. While the Base model maintains near-zero hallucination rates (5.4%), FT-A exhibits 82.5% overt hallucination and FT-A+B+C reaches 98%. Contrary to intuition, numeracy supervision amplifies rather than mitigates hallucination across all levels. We identify template injection—the insertion of memorized canonical values regardless of input content—as a primary hallucination mechanism in fine-tuned models. These findings demonstrate that numerical hallucination in financial summarization is driven by the degradation of numerical restraint through domain adaptation, not by insufficient numerical reasoning. We recommend that evaluation protocols assess hallucination across all detectability levels and that deployment practices include explicit mechanisms for grounding-aware generation or abstention.
Read the full abstract

Mathematical Physics

Derived Associative Algebras and Cyclic Cohomology of Q-manifolds

Loday's derived product is revisited in the setting of Q-manifolds, i.e., supermanifolds equipped with an odd vector field that `squares to zero'. We interpret the Grassmann odd product as a standard product upon shifting the grading, which implies the existence of a derived (noncommutative) associative $\mathbb{Z}_2$-graded algebra associated with any Q-manifold. We apply Connes' cyclic cohomology to the derived associative algebra, giving a new cohomology on Q-manifolds distinct from the standard cohomology: we refer to this as the derived cyclic cohomology of a Q-manifold. The derived cyclic cocycles are interpreted as classically BRST-invariant functionals within the BV–BFV–BRST formalism or generalised Ruelle–Sullivan currents when applied to regular foliations via their foliation Lie algebroids.
Read the full abstract

Artificial Intelligence

MedFlow: Class-Aware Multi-Scale Generation for Medical Time-Series Synthesis

Synthetic medical time-series generation can alleviate data scarcity and support the development of reliable clinical prediction models. However, existing methods mainly focus on matching the overall distribution and temporal dynamics of real data, which does not necessarily ensure strong downstream utility on imbalanced medical datasets. Clinically informative patterns often occur at heterogeneous temporal scales, while rare minority-class characteristics can be obscured by dominant population patterns. To address these challenges, we propose MedFlow, a class-aware multi-scale flow matching framework for medical time-series synthesis. MedFlow employs a vector-quantized multi-scale tokenizer to represent medical sequences at complementary temporal resolutions, capturing both coarse clinical trends and fine-grained dynamics. We further introduce Token Marginal Guidance, which incorporates class-conditional token statistics directly into the flow matching process to steer generation toward class-specific regions of the learned tokens. This mechanism strengthens minority-class patterns, while preserving the global and tail distributions of real data. Experiments on four public datasets covering electronic health records, EEG, and ECG signals demonstrate that MedFlow consistently outperforms recent state-of-the-art diffusion-based baselines across downstream prediction tasks. On average, it improves AUPRC by 5.8%, reduces Context-FID by 88.6%, and achieves 3.8$\times$ higher sampling throughput.
Read the full abstract

All authors Yanhao Huang, Shibo Feng, Wanjin Feng, Peilin Zhao, Chunyan Miao

Artificial Intelligence

Hierarchical Possession-Aware Graph Pointer Network for Pass Receiver Selection

Pass receiver selection is a fundamental task in football analytics, aiming to predict the intended receiver under a given game state. This task is challenging with event-centered freeze-frame observations, a broadcast-like setting that provides only partial and variable player visibility without complete trajectories or stable player identities. The model must therefore reason over anonymous visible candidates, opponent pressure, and recent context under partial observation. To address this setting, we propose a Hierarchical Possession-aware Graph Pointer Network (HPGPN), which formulates pass receiver selection as variable-size candidate prediction over visible teammates. HPGPN jointly models current player interactions, local event context, and possession-level temporal dynamics. It represents the current pass situation with a graph, incorporates fixed event context, and uses dynamic possession history to capture how the attacking sequence evolves. Candidate representations are refined hierarchically by integrating spatial, contextual, and historical evidence, and a glimpse pointer head scores the receiver candidates. Experiments on public football event and freeze-frame data show that HPGPN improves pass receiver selection performance. Ablation studies demonstrate the effectiveness of graph-based interaction modeling, fixed event context, and dual-branch dynamic possession-history modeling.
Read the full abstract

All authors Jingyi Wang, Da Li, Kaixin Wang, Zhangqin Huang

Computer Vision

Linguistic Trajectory Encoding for Efficient Long-Horizon Spatial Memory in Embodied Agents

Embodied agents performing long-horizon tasks require a memory representation in which the state transitions of dynamic objects remain queryable in natural language across hours-to-days observation horizons. Existing systems either drop fine-grained motion (clip-level video-language embeddings), keep it only as raw coordinates (geometric SLAM), or organise it around immediate task context (agent working memories). None of them gives the agent a per-object timeline whose state transitions are themselves queryable in language. Our key contribution is Linguistic Trajectory Encoding (LTE), which compresses dynamic object motion histories via a hybrid representation combining natural language descriptions, sparse spatial anchors, and visual anchors. LTE adapts compression to motion complexity by anchoring periods without reliable observations to the last seen location, while representing motion with geometric waypoints and linguistic descriptions to preserve accuracy. To evaluate these capabilities across extended time horizons, we construct the Spatial Memory Benchmark (SMB) from EgoLife multi-day recordings, targeting capabilities absent in existing benchmarks: semantic trajectory retrieval and long-horizon object retrieval. On SMB, the LTE-based system achieves $45.3\%$ success in semantic trajectory retrieval and $48.7\%$ in long-horizon object retrieval, outperforming structured-memory and VLM baselines (best prior: $31.9\%$ and $34.4\%$). LTE achieves trajectory compression by factors of $8.7\times$ to $26.1\times$ with sub-second query latency on $24$\,h video. On Ego4D natural-language queries, the system reaches $28.75\%$ / $55.10\%$ R@1/R@5, $+15.80$ / $+31.30$ pts over EgoVLPv2.
Read the full abstract

All authors Tianyidan Xie, Shenyi Wang, Qiang Tang, Mingjie Wang, Zhicheng Qiu, Xuanfu Li, Zhan Xu, Jian Yang, Lanjun Wang, Zili Yi

Artificial Intelligence

Whose record is this? Diagnosing and authorizing record use in personalized multimodal models

Contextualized visual personalization can retrieve a true record yet apply it to the wrong visual subject. We formalize when a record may condition an answer as record authorization: subject presence ($P$), record-edge validity ($E$), and answer support ($S$) must all hold. We call violations visual memory misbinding (VMM). We construct RecordAuth-Diag, a 3,690-case matched diagnostic suite that changes one image–record edge while holding the query, question, record text, and image multiset fixed. Card removal and nonce relabeling attribute these failures to supplied records. Raw-bank failures span Qwen-, Phi-, and Gemma-family interfaces: Gemma-3-4B-IT reaches 63.69% local unauthorized use at 25.75% clean recall. CoViP remains at 26.02%, versus 22.49% for its Qwen backbone at similar clean recall. Typed pre-generation authorization reduces Qwen card exposure on RecordAuth-Diag from 43.63% to 3.06%, while positive recall changes from 86.26% to 60.90%. Full $P\wedge E\wedge S$ validation uses 560 localized DAVIS cases: top-1 relevance and typed authorization have comparable release (28.93% and 28.39%) but 6.79% and 0.89% unsafe release, respectively. Of the 33 additional unsafe cases removed, 27 are support, 4 edge, 2 clean, and 0 boundary cases. Thus the observed increment is an $E\wedge S$ decision dominated by support, not an edge check alone. Appearance supplies $E$ evidence only conditional on $P$; authenticated subject tokens instantiate the missing presence witness as a sufficiency control. The claims concern the evaluated contracts, not natural prevalence, consent, or visual identity
Read the full abstract

All authors Xinyu Mao, Junsi Li, Chenyang Liu, Haoji Zhang, Ming Sun

Computer Vision

Intrinsic Temporal Adaptation of CLIP for Partially Relevant Video Retrieval

Partially Relevant Video Retrieval (PRVR) aims to retrieve untrimmed videos that contain moments relevant to a text query. Since the target moment occupies only a portion of the video, PRVR requires retrieval based on fine-grained understanding beyond coarse video-level matching. However, existing methods often rely on frozen CLIP frame features, which lack temporal understanding. Even with recent progress in parameter-efficient CLIP adaptation, video-level predictions can still be supported by imprecise frame-level evidence. In this paper, we propose an Intrinsic Temporal Adaptation (ITA) framework for PRVR. First, our Backbone-Internal Temporal Adaptation allows the last few visual transformer layers to attend over groups of neighboring frames. This provides temporally aware frame embeddings while keeping CLIP frozen and training only adaptation parameters. Second, we introduce Affinity-Weighted Gradient Propagation to address the weakly supervised nature of PRVR, softly aggregating top-$k$ frames based on text-frame affinities and propagating learning signals to multiple query-relevant frames. Our method achieves state-of-the-art performance on PRVR benchmarks, demonstrates robust cross-dataset transfer, and retrieves substantially more accurate frame-level evidence within ground-truth query-relevant moments. Our code is available at github.com/hynnsk/ITA.
Read the full abstract

All authors Hyun Seok Seong, Woojin Jun, SuBeen Lee, Jae-Pil Heo

Robotics

HaptiNet: Networked Haptic Robots Enable Physical Co-presence in Geographically-Unconstrained Rehabilitation

Cooperative rehabilitation enhances engagement, task performance, and social-motor interaction, yet it demands physical co-presence: users must transmit forces, coordinate movements, and infer intent through haptic contact. Telerehabilitation promises to expand access for patients constrained by distance, mobility, or clinical disparities, yet current techniques remain predominantly audiovisual while leaving users haptically and physically isolated. Here, we introduce HaptiNet, a networked haptic robotic system enabling physical co-presence for geographically distributed users via force-mediated interaction. Each robotic terminal features a low-inertia, long-stroke design with high force-feedback capacity, tailored for haptic rendering in upper-limb training. Building on these terminals, HaptiNet creates a distributed haptic network with an imitation-learning-based delay compensator, enabling users to physically perceive and coordinate with one another over distance. We validated HaptiNet in 284 healthy participants and 111 patients with neurological impairments across progressively realistic settings, including laboratory tests, cross-city deployments, and clinical applications. HaptiNet preserved task-level force rendering consistency across single-user and multi-user scenarios. Compared with solo and visual cooperative training, haptic cooperation improved task performance by 24% and 22%, respectively, while also boosting engagement and interpersonal motor synchrony. Across three intercity links totaling approximately 4,000 km, HaptiNet maintained stable haptic interaction among patients with neurological impairments, producing a 3.87-fold greater baseline-to-training score improvement and a 106% higher patient-applied effort over the solo condition.
Read the full abstract

All authors Chenyang Sun, Mingjie Dong, Haodong Deng, Yudong Liu, Yi-Feng Chen, Jun Lin, Changlong Huang, Jie Guo, Yantong Liu, Yang Liu, Yuzhou Lin, Jianjun Long, Zheng Xing, Sining Zhao, Xuemin Zhang, Zhiyong Wang, Zhenhong Li, Dongrui Wu, Honghai Liu, Jian S. Dai, Mingming Zhang

Materials Science

Field-Driven Coupled Magnon–Phonon–Electron Relaxation in Magnetite Using Steepest-Entropy-Ascent Quantum Thermodynamic Formalism

A field-driven steepest-entropy-ascent quantum thermodynamic (SEAQT) formulation is developed for longitudinal nonequilibrium relaxation in magnetite (Fe$_3$O$_4$) with coupled electron, phonon, and magnon populations. Material-specific excitation spectra define the thermodynamic state space, while one relaxation parameter for each population sets its kinetic scale. A longitudinal magnetic field shifts the dressed magnon eigenenergies while the occupation basis remains fixed; irreversible redistribution conserves instantaneous energy and electron number while allowing the magnon population to vary. The formulation yields nonequilibrium subsystem temperatures, entropy production, magnetic-work identities, and a coupled small-signal susceptibility incorporating energy-conservation feedback among all three populations; the one-pole Debye response appears only as a limiting case. Numerical results under sinusoidal driving show a transition from nearly quasistatic behavior to frequency-dependent lag, finite-amplitude departure from the linear-response ellipse, and increasing higher-harmonic content. Relaxational work per cycle increases strongly with field amplitude and frequency, while the complex susceptibility is broader and shifted relative to a Debye reference. Entropy production remains positive, and the electron, phonon, and magnon temperatures show distinct excursions followed by secular heating when positive magnetic work is retained without heat rejection. The calculated work represents longitudinal magnon quasiparticle relaxation in a homogeneous single-domain model, not the total core loss of a finite ferrite specimen.
Read the full abstract

All authors Deepak Dhariwal, William T. Reynolds, Jr., Michael R. von Spakovsky

Machine Learning

How Faithful Is Attribution for Sales Forecasting? A Counterfactual Study

Deep models for sales forecasting, such as WaveNet-style dilated convolutional networks, are accurate but opaque: when a single model predicts sales for one of many series, it offers no account of why. We add a post-hoc, architecture-agnostic counterfactual interpretability layer to a multi-series WaveNet forecaster trained on the full Corporacion Favorita grocery dataset (174,685 series over 1,688 days). The method decomposes each forecast into contributions that sum exactly to the predicted value, avoiding the allocation artifacts we observed with additive SHAP-style attribution. We evaluate faithfulness with a deletion/insertion protocol and find a statistically significant effect on both tests (deletion gap 0.22, p<0.001; insertion gap 0.27, p<0.01; robust across five background-sampling seeds), establishing that the attributions reflect genuine model behavior rather than plausible-looking artifacts. We then characterize, honestly, where attribution is and is not informative: reliance on the promotion signal is heterogeneous across series (median ratio approximately 1.0, with roughly 20% of series showing a strong effect), and the model captures the shape of the weekly sales cycle (day-of-week r=0.78) while systematically under-predicting its amplitude. Our contribution is not improved accuracy but an interpretability layer with a rigorous faithfulness evaluation and a candid account of its limits.
Read the full abstract

Algebraic Geometry

Superconnections, descent, and monodromy on transversely holomorphic foliations

Let $X$ carry a transversely holomorphic foliation, equivalently an elliptic involutive structure $V\subset T_{\mathbb C}X$, and let ${\mathcal O}_V$ be its sheaf of leafwise-constant, transversely holomorphic functions. We construct a finite superconnection model for the derived category of coherent ${\mathcal O}_V$-modules. The key input is a mixed local reduction for finite Maurer–Cartan objects over the mixed de Rham–Dolbeault dga $(\wedge^\bullet V^\vee,d_V)$: a multiplicative homotopy contracts the real directions, after which Block's Dolbeault gauge theorem removes the positive transverse form degrees. For compact $X$, this gives an exact equivalence between the homotopy category of bounded finite-rank flat $V$-superconnections and $D^b_{\mathrm{coh}}(X,{\mathcal O}_V)$, interpolating between the de Rham and Dolbeault realizations. We prove locally finite Čech descent under necessary uniform amplitude and rank bounds, identify the coherent heart with equivariant coherent analytic sheaves on a transverse monodromy groupoid, and characterize descent to ordinary holonomy. For a holomorphic suspension, we identify the full superconnection category, up to Morita equivalence, with the homotopy fixed points of the Dolbeault category of the transversal and derive an equivariant Ext spectral sequence. Examples on $S^1$ and $S^2$ delimit when ordinary monodromy $1$-groupoids can recover the derived category.
Read the full abstract

K-Theory and Homology

Products in $KK$- and $E$-theory

In this note, we give an explicit description of countable products in $KK$- and $E$-theory and provide several applications.

Optimization and Control

Policy Iteration for Domain Randomized Linear Quadratic Systems

In this work, we study policy optimization under domain randomization for linear quadratic control, focusing on learning a single state-feedback controller that minimizes the average cost across systems with uncertain dynamics. We propose a policy iteration algorithm with a step-size rule that preserves stability across all sampled systems at each iteration. We show that the method yields monotonic improvement of the sample-average objective and that a stabilizing step size always exists. Under standard smoothness assumptions, the iterates converge subsequentially to stationary points, and under a gradient-dominance condition, we obtain a global linear convergence rate.
Read the full abstract