arXiv:2605.26858v1 Announce Type: new
Abstract: There has been a proliferation of media reports about so-called AI psychosis in the last year. Not surprisingly, this has prompted growing academic work on the ways in which AI chatbots such as ChatGPT, Claude, and Replika might aggravate or even induce psychosis, typically understood in terms of users acquiring or maintaining delusional beliefs. Our paper consists of two parts. First, we provide a number of reasons to be sceptical about understanding 'AI psychosis' as a novel psychiatric category. We argue that many of the purportedly new phenomena are better understood through Stompe et al.'s (2003) metaphor of 'old wine in new bottles' and highlight conceptual, nosological, clinical, and social risks associated with the uncritical adoption of this terminology. Second, we develop a positive phenomenological account of what may nevertheless be at stake in sustained human-AI interaction. Rather than focusing primarily on whether AI systems induce, amplify, or sediment delusional beliefs, we examine how conversational AI may participate in transforming a person's lived experience of reality itself. We claim that the sycophantic and pseudo-intersubjective nature of AI could lead to what we call "existential drift", whereby individuals may continue to feel rooted in a shared reality through their interactions with AI, while actually becoming entrenched in increasingly private and subjective worlds.
Science Journals
arXiv:2605.26680v1 Announce Type: new
Abstract: Recent video multimodal large language models (MLLMs) increasingly couple step-by-step reasoning with on-demand visual evidence retrieval, allowing models to revisit relevant video segments during inference. However, two structural gaps remain in existing thinking-with-video systems. (i) Sampling density is not a learnable decision: existing methods may let the model decide where to look, but the per-window frame rate is largely fixed. As a result, fine-grained evidence is often recovered through repeated retrieval calls, which increases inference context length and training difficulty. (ii) Retrieval and answer generation are usually optimized with a single trajectory-level advantage, so the "where to look" tokens and the "how to answer" tokens receive the same credit even when one is correct and the other is not. To address these gaps, we present DynFrame, a framework that emits the temporal window and the sampling density as native tokens within a single autoregressive pass. This learnable span-density retrieval enables acquiring multi-granularity evidence with a single retrieval step. Based on the above tokenized retrieval interface, we further introduce Segment-Decoupled GRPO (SD-GRPO), which splits each rollout at the retrieval boundary and assigns role-specific token-level advantages, separately crediting the sampling decision and the answer. Trained on the curated DM-CoT-74k and DM-RL-45k, DynFrame-4B is competitive with strong 7B-8B baselines across six benchmarks (NExT-GQA, Charades-STA, ActivityNet-MR, Video-MME, MLVU, LVBench), and DynFrame-8B sets new state-of-the-art on most metrics. Code is available at https://github.com/zhangguanghao523/DynFrame.
arXiv:2512.22157v2 Announce Type: replace
Abstract: This paper presents a matrix formulation of the scalar laws of radiative transfer. The method applies to coupled mixed boundary condition problems on general domains. Participating media can range from transparent to absorbing, emitting, and scattering, with boundaries ranging from absorbing to reflecting. Given a non-dimensional first-interaction exchange factor matrix $\mathbf{F}$, the formulation partitions $\mathbf{F}$ into a single-step absorption matrix and a single-step reflection-scattering matrix via Hadamard products with a column-constant matrix of reflection-scattering coefficients. The resulting linear system encodes the radiative energy balance for arbitrary combinations of prescribed temperatures and prescribed source terms, with a proven non-singularity result for the mixed-boundary system. The method is shown to admit a unique non-negative solution for non-negative source terms whenever the maximum reflection-scattering coefficient is strictly less than unity, with unconditional energy conservation to machine precision. Validation is provided symbolically against the textbook closed-form solutions for infinite parallel plates and concentric cylinders, and numerically against the diffusion approximation in the high-extinction limit and against the results of Crosbie and Schrenker for pure and partial scattering cases. A comparison with Noble's matrix formulation of Hottel's zonal method reveals a discrepancy in that classical approach, not previously identified to the author's knowledge; the proposed formulation avoids this discrepancy. The method requires a single linear solve whose sparsity inherits from that of $\mathbf{F}$, making it applicable to medium-scale dense problems and to large-scale sparse problems with high extinction.
arXiv:2509.18919v2 Announce Type: replace
Abstract: The pretraining-finetuning paradigm is a crucial strategy in metallic surface defect detection for mitigating the challenges posed by data scarcity. However, its implementation presents a critical dilemma. Pretraining on natural image datasets such as ImageNet, faces a significant domain gap. Meanwhile, naive self-supervised pretraining on in-domain industrial data is often ineffective due to the inability of existing learning objectives to distinguish subtle defect patterns from complex background noise and textures. To resolve this, we introduce Anomaly-Guided Self-Supervised Pretraining (AGSSP), a novel paradigm that explicitly guides representation learning through anomaly priors. AGSSP employs a two-stage framework: (1) it first pretrains the model's backbone by distilling knowledge from anomaly maps, encouraging the network to capture defect-salient features; (2) it then pretrains the detector using pseudo-defect boxes derived from these maps, aligning it with localization tasks. To enable this, we develop a knowledge-enhanced method to generate high-quality anomaly maps and collect a large-scale industrial dataset of 120,000 images. Additionally, we present two small-scale, pixel-level labeled metallic surface defect datasets for validation. Extensive experiments demonstrate that AGSSP consistently enhances performance across various settings, achieving up to a 10\% improvement in mAP@0.5 and 11.4\% in mAP@0.5:0.95 compared to ImageNet-based models. All code, pretrained models, and datasets are publicly available at https://clovermini.github.io/AGSSP-Dev/.
arXiv:2605.26952v1 Announce Type: new
Abstract: Agentic reinforcement learning (RL) has proven effective for training LLM-based agents with external tool-use capabilities. However, we identify that agentic RL training induces increasing redundant tool calls and blurs the model's intrinsic knowledge boundary, where the model fails to distinguish when tools are needed versus when parametric knowledge suffices. Existing solutions based on reward shaping create coarse-grained optimization targets that tend to incentivize indiscriminate tool-call suppression, leading to reward hacking. In this paper, we propose AKBE (Agentic Knowledge Boundary Enhancement), an on-policy method that dynamically probes the model's intrinsic knowledge boundary through dual-path (with-tool and no-tool) rollouts during training. We define the knowledge boundary as the per-instance determination of whether tools are required and the minimum tool calls necessary. By comparing correctness across paths, AKBE categorizes trajectories and constructs targeted supervisory signals that guide efficient tool-use patterns for each question. These signals are integrated seamlessly into the agentic RL training loop. Experiments on seven QA benchmarks demonstrate that AKBE improves task accuracy by +1.85 on average and reduces tool calls by 18% over standard agentic RL, yielding 25% higher tool productivity without any accuracy-efficiency trade-off. Further analysis suggests its plug-and-play compatibility across different RL algorithms and the mechanism of each signal category. Our code is available at https://github.com/CuSO4-Chen/AKBE.
arXiv:2605.26949v1 Announce Type: new
Abstract: 3D shape completion from partial scans remains challenging for unseen categories and noisy real-world observations, where geometry alone is often insufficient for inferring missing structure. We present DinoComplete, a deterministic and efficient shape completion framework that augments geometric reconstruction with voxel-aligned semantic priors distilled from DINO features. First, we construct multi-view DINO feature volumes aligned with ShapeNet data and train a student network to predict dense semantic features directly from incomplete shapes. These predicted features capture global structure and part-aware semantic context while remaining aligned with the underlying geometry. We then integrate these distilled features into a completion network, where geometric and semantic voxel representations are fused through voxel state-space modeling. To enable efficient long-range reasoning without sacrificing resolution, we introduce a multi-scale voxel Mamba module that refines the fused features by combining full-grid and chunk-wise sequence modeling. Experiments on unseen ShapeNet categories and ScanNet objects show that DinoComplete achieves stronger completion quality than prior deterministic and generative based completion methods while using fewer parameters, requiring lower memory, and achieving faster inference. Our results demonstrate that distilling semantic priors from visual foundation models improves generalization and robustness in 3D shape completion.
arXiv:2605.27268v1 Announce Type: new
Abstract: Modern Large Language Models (LLMs) are often criticized for producing repetitive and homogeneous text, despite possessing vast latent vocabularies. While previous research has focused on model knowledge and training data, we investigate the role of decoding mechanics in suppressing linguistic diversity. We introduce the Word Coverage Score (WCS), a metric that quantifies the extent to which contextually appropriate human vocabulary is mathematically pruned by standard sampling filters (e.g., Top-$p$, Top-$k$, and Min-$p$). Rather than assessing static knowledge, the WCS measures the lexical survival rate of low-frequency, high-information human words as a function of sampling parameters. By auditing open-weight models on human-authored corpus fragments, we identify which logical lexical choices are rendered unreachable by the decoder, even when they reside within the probability space. Our results provide quantitative evidence that industry-standard sampling defaults act as unintended censorship mechanisms, smoothing the unique textures of human expression into a homogenized discourse. The WCS offers a rigorous framework for optimizing the trade-off between text coherence and lexical richness, providing a diagnostic tool for preserving the diversity of human language in generative models.
arXiv:2605.26445v1 Announce Type: new
Abstract: Physicians learn primarily about illicit drugs from clinical overdose cases, limiting their understanding of real-world usage. Meanwhile, drug users share first-hand experiences online, offering insights into dosage and effects of drugs. To bridge this gap, we introduce ReDose (REddit Drug DOSe and Effect), a dataset of 6,435 Reddit posts on substance use. A board-certified toxicologist primarily annotated both the training and test sets, while two medical science students contributed to the test set, labeling DRUG, DOSE, and EFFECT entities. We benchmarked 6,267 annotations using BERT-based, large language model (LLM)-based, and Retrieval-Augmented Generation (RAG) models. BiomedBERT achieved an F1-score of 0.843 for DRUG, while Llama-3 70B outperformed GPT-4 (F1 = 0.79 vs. 0.72). EFFECT extraction remains challenging, with GPT-4 achieving a recall of 0.41. ReDose captures patient-curated narratives to advance medical data extraction from social media.
arXiv:2512.13268v2 Announce Type: replace
Abstract: High-performance computing (HPC) systems consume enormous amounts of energy, with idle nodes as a major source of energy waste. Powering down idle nodes can mitigate this problem, but long boot/shutdown delays can introduce significant queueing penalties if transitions are poorly timed. To address this trade-off, we present SPARS, a reinforcement learning-enabled simulator for power management in HPC job scheduling. SPARS integrates job scheduling and node power-state management within a discrete-event simulation framework. It supports traditional scheduling policies such as First Come First Serve and EASY Backfilling, along with enhanced variants that employ reinforcement learning agents to dynamically decide when nodes should be powered on or off. Users can configure workloads and platforms in JSON format, specifying job arrivals, execution times, node power models, and transition delays. The simulator records comprehensive metrics-including energy usage, wasted power, job waiting times, and node utilization-and provides Gantt chart visualizations to analyze scheduling dynamics and power transitions. Unlike widely used Batsim-based frameworks that rely on heavy inter-process communication, SPARS provides lightweight event handling and consistent simulation results, making experiments easier to reproduce and extend. Its modular design allows new scheduling heuristics or learning algorithms to be integrated with minimal effort. By providing a flexible, reproducible, and extensible platform, SPARS enables researchers and practitioners to systematically evaluate power-aware scheduling strategies, explore the trade-offs between energy efficiency and performance, and accelerate the development of sustainable HPC operations.
arXiv:2512.21602v3 Announce Type: replace
Abstract: Every year, millions of patients pass through emergency departments and intensive care units, where clinicians must make high-stakes decisions under time pressure and uncertainty. Machine learning could support prediction of deterioration, triage, and rare critical outcomes, but clinical data are often severely imbalanced, biasing models toward majority classes and reducing predictive performance. Developing robust and efficient models for imbalanced clinical tabular data therefore remains an important challenge.
We evaluated six model families on imbalanced tabular data from the MIMIC-IV-ED and eICU databases: Decision Tree, Random Forest, XGBoost, TabNet, TabICL, and TabPFN v2.6. Trainable models were optimized using Bayesian hyperparameter tuning, while foundation models were evaluated in their pretrained inference regime without task-specific reweighting. Models were assessed using Macro F1-score, robustness to increasing imbalance, and computational scalability across seven clinical prediction tasks.
Results differed across datasets. On MIMIC-IV-ED, TabPFN v2.6 and TabICL achieved the strongest average Macro F1 ranks, with XGBoost remaining competitive. On eICU, XGBoost consistently performed best, followed by other tree-based methods, while foundation models achieved intermediate performance. Across both datasets, TabNet showed the largest degradation under increasing imbalance and the highest computational cost. Training-time analysis showed that tree-based methods scaled most favorably with dataset size, while foundation models offered low per-task adaptation cost.
These findings suggest that no single model family dominates across all clinical settings. However, tabular foundation models are narrowing the performance gap with strong classical baselines while offering a distinct efficiency-performance trade-off that may benefit resource-constrained clinical environments.
arXiv:2605.26937v1 Announce Type: new
Abstract: Parametric knowledge in large language models (LLMs) is a cornerstone of their success, yet remains poorly understood. Existing knowledge benchmarks typically rely on predefined questions (e.g., "What is the birth date of M.L. King?"), evaluating only knowledge that benchmark designers explicitly choose to query, a problematic availability bias.
In this paper, we introduce open knowledge evaluation, a new paradigm for LLM knowledge benchmarking. Instead of asking narrow questions, it evaluates models on the knowledge they choose to surface in response to open-ended elicitation prompts (e.g., "Tell me everything you know about M.L. King"). This shifts the focus from predefined answer retrieval toward characterizing the knowledge models naturally express.
We instantiate this paradigm with BeQu (Beyond Questions), a benchmark of 10,000 entities paired with reference corpora for statement verification. Using BeQu, we evaluate a broad range of language models and analyze the effects of reasoning effort, model scale, prompt format, and knowledge domain. Data and leaderboard are available on this work's GitHub repository and at the benchmark's website.
arXiv:2605.20988v2 Announce Type: replace
Abstract: We study transformers' generalization behavior on boolean domains from the perspective of the Fourier spectra of their target functions. In contrast to prior work (Edelman et al., 2022; Trauger & Tosh, 2024), which derived generalization bounds from Rademacher complexity, we investigate the feasibility of obtaining generalization bounds via PAC-Bayes theory. We show that sparse spectra concentrated on low-degree components enable low-sharpness constructions with good generalization properties. Our idea is to show the existence of flat minima implementing any boolean function of sparsity no greater than the context length, and then apply a PAC-Bayes bound to an idealized low-sharpness learner, resulting in a non-vacuous generalization bound. We use this to give a formal account of why chain-of-thought improves generalization for high-degree target functions, and show that the complexity parameters in our bound can be efficiently estimated via property testing. We evaluate predictions empirically and conduct a mechanistic interpretability study to support the realism of our theoretical construction in real transformers.
arXiv:2605.19908v2 Announce Type: replace
Abstract: Authorship attribution models fine-tuned with the same pretrained encoder, data, and loss can differ four-fold in performance depending only on their scoring mechanism. We use mechanistic interpretability tools to explain this gap. Stylistic features such as word length, punctuation density, and function-word frequency are similarly available at every layer in every model we probe, including an off-the-shelf control encoder, suggesting that the gap is not explained by their linear readability. Instead, causal intervention shows that the scorer appears to determine where the encoder consolidates authorship signal. Mean pooling forces consolidation by early to mid layers, while late interaction defers it to later layers. We further derive this difference from the gradient structure of each scorer, and training dynamics reveal distinct learning trajectories that follow from that difference.
arXiv:2605.26878v1 Announce Type: new
Abstract: Multi-stakeholder tasks require one output to satisfy users with conflicting preferences. Holistic LLM judges conflate utility estimation and utility aggregation, yielding unstable implicit weights. We show empirically and theoretically that this aggregation-specific \emph{weighting noise} can create large score shifts when stakeholder satisfaction is dispersed; in our experiments, these weight-induced shifts also increase with stakeholder count. We propose \textsc{DecompR}: counterfactual-calibrated weights are fixed from query structure before candidate scoring, while per-role utilities are estimated independently, removing candidate-dependent weight drift and reducing estimation noise.
arXiv:2605.25480v2 Announce Type: replace
Abstract: LLM agents require retrieval to behave less like one-shot context fetching and more like reasoning: searching, reading, traversing, and deciding when evidence is sufficient. Yet current Retrieval-Augmented Generation (RAG) systems organize external knowledge as flat chunks retrieved by embedding similarity, exposing a retrieval-as-lookup interface ill-suited to iterative reasoning agents. We propose LLM-Wiki, an agent-native retrieval system that operationalizes the Retrieval-as-Reasoning paradigm by treating external knowledge as a compilable, composable, and self-evolving structure rather than a static retrieval index. LLM-Wiki compiles documents into structured Wiki pages with bidirectional links, exposes search, read, and link-following operations through standard tool-calling interfaces, and introduces an Error Book for persistent structural and semantic self-correction. LLM-Wiki achieves state-of-the-art results on HotpotQA, MuSiQue, and 2WikiMultiHopQA, outperforming HippoRAG 2, LightRAG, and GraphRAG by 2.0-8.1 F1 points. On AuthTrace, LLM-Wiki achieves the best overall accuracy, with especially strong gains on multi-document structured queries, confirming that compilation-based retrieval generalizes beyond chain-style multi-hop reasoning.
arXiv:2605.20251v4 Announce Type: replace
Abstract: Existing benchmarks for LLM coding agents primarily evaluate final outcomes. While useful for measuring overall capability, these metrics provide limited visibility and often miss defects that arise during execution. We present ProcCtrlBench, a benchmark for execution-process evaluation in LLM coding agents. ProcCtrlBench organizes recurrent execution defects into a reusable ontology covering 11 defect types in 4 categories, and evaluates agent trajectories through standardized process evidence rather than final outcomes alone. To support comparison across heterogeneous agents, ProcCtrlBench standardizes raw logs into a unified trajectory representation and reports calibrated scorecards over process-level findings. In addition, ProcCtrlBench uses control preservation as a way to quantify execution-process quality, capturing whether execution remains interpretable, interruptible, correctable, reversible, and able to hand back authority when needed. We evaluate ProcCtrlBench on 200 cases sampled from three benchmarks: AndroidBench, TerminalBench, and SWE-bench-Verified. Results show that ProcCtrlBench can be instantiated with useful reliability, provides more stable semantics than direct thresholding, and reveals meaningful differences in execution quality that are often overlooked by conventional outcome-based evaluation.
arXiv:2605.26341v1 Announce Type: new
Abstract: Physics-informed machine learning (PIML) integrates mechanistic knowledge, typically in the form of partial differential equations (PDE), into data-driven models. Despite strong empirical performance, its statistical generalisation properties remain poorly understood, particularly in the regression setting with unbounded losses. Existing analyses rely on approximation or stability arguments and do not fully capture how physical structure influences generalisation from finite data. In this work, we develop a PAC-Bayesian framework for PIML that provides high-probability generalisation guarantees in the presence of unbounded losses. We adopt a multi-task perspective that jointly treats data fidelity, PDE residuals, initial and boundary conditions, avoiding the looseness induced by standard union-bound approaches. Our analysis leverages the structure of physics-informed objectives to derive novel bounds where the complexity scales with input-gradient norms of the losses, revealing a direct link between physical regularity and generalisation. We instantiate this framework under Sobolev and Poincar\'e-type assumptions, yielding two classes of bounds that trade off statistical complexity and smoothness in different regimes. Building on these results, we propose a self-bounding-aware learning algorithm that directly optimises tractable surrogates of the derived bounds, along with a practical procedure to estimate the associated constants in realistic settings. Empirical evaluations on standard PDE benchmarks demonstrate that our bounds are non-vacuous, significantly tighter than union-bound baselines, and can be effectively minimised during training. Overall, our results provide a principled statistical foundation for the generalisation of physics-informed models.
arXiv:2605.26699v1 Announce Type: new
Abstract: Anthropogenic underwater noise from offshore wind turbines is a growing environmental concern, particularly with the large-scale deployment of bottom-fixed and floating devices. This study presents a physics-based vibroacoustic framework to predict operational underwater noise emissions from offshore wind turbines and compares monopile-supported and floating configurations for a 10 MW turbine. The methodology combines time-domain aero-hydro-servo-elastic simulations with a frequency-domain acoustic formulation based on equivalent dipole sources and Green's function solutions, accounting for underwater confinement between the free surface and seabed through the method of images. Results show that floating configurations exhibit enhanced low-frequency acoustic emissions, producing up to 15% higher OASPL than the monopile structures under equivalent water depths for frequencies below 10 Hz due to additional rigid-body motions, while monopile structures radiate more efficiently at higher frequencies associated with drivetrain excitations. Significant differences in the spatial distribution and directivity of the radiated sound field are also observed, with floating platforms displaying more complex three-dimensional radiation patterns and stronger direction-dependent variations, reaching approximately 20-25 dB in the 100-1000 Hz band, compared with the smoother and nearly axisymmetric response of monopile configurations. Water depth strongly influences propagation regimes and overall sound levels, with shallow-water floating configurations showing variations of up to 7% in OASPL relative to deep-water cases. The proposed framework enables quantification of vibro-acoustic noise and provides a predictive tool for assessing underwater acoustic impacts during the design phase, supporting environmentally informed offshore wind turbine design and future regulatory and monitoring strategies.
arXiv:2507.00637v3 Announce Type: replace
Abstract: Cyberattacks on enterprise networks exploit complex dependencies among infrastructure, services, and applications, which challenge traditional analysis methods that focus on attack paths or network topology in isolation. In this study, we introduce a novel probabilistic multilayer modelling framework, based on influence propagation in networks, that integrates attack graphs with the communication network topology, enabling a service-centric impact analysis of cyberattacks. Our method captures both the vulnerability exploitability and network connectivity, allowing us to assess the likelihood of attack propagation and cumulative impacts across interconnected services. By integrating standard vulnerability metrics (such as CVSS) with the network-level connectivity probabilities, the framework provides a cohesive view of the dynamics of cyberattacks. We validate this approach using a realistic case study of an enterprise network, demonstrating its ability to determine critical nodes, vulnerabilities, and service dependencies that significantly influence attack outcomes. Our findings show that integrating network and attack graph perspectives offers more actionable insights into risk assessment and mitigation planning, advancing the analysis of cyberattacks in complex networked environments.
arXiv:2604.23541v3 Announce Type: replace-cross
Abstract: This paper presents a symmetric unified transport (UT) compact model for metal-oxide-semiconductor field-effect transistors (MOSFETs) that bridges drift-diffusion (DD) and ballistic transport (BT) regimes. The proposed model self consistently accounts for both current and charge across the DD-BT transition. Quantum capacitance and carrier transport are incorporated into the charge density formulation. Drain side velocity saturation and the source side thermal velocity limit are unified within a single framework using a physically motivated high field scattering length, enabling accurate modeling from DD square law behavior to the ballistic limit. In addition, a physical channel charge and capacitance model is developed to capture capacitance reduction in the quasi-ballistic regime, which is not considered in standard compact models. The model is verified using theoretical analysis and experimental data from MOSFETs with multiple channel lengths, achieving accurate fitting using only physically motivated model parameters. The formulation is continuous and symmetric, and it passes both DC and AC symmetry tests.
arXiv:2601.09358v2 Announce Type: replace
Abstract: Large regions of giant planets are thought to possess unstable thermal gradients stabilised by gradients in heavy-element composition. The fluid can then develop semi-convection, a double-diffusive instability driven by the unequal molecular diffusivities of heat and composition. While previous studies have focus mainly on local Cartesian models, we investigate semi-convection in rotating spherical shells, the geometry relevant to planetary interiors, using direct numerical simulations In a first nonlinear phase, the flow spontaneously forms concentric density staircases composed of well-mixed layers separated by thin, strongly stratified interfaces. We propose scalings for both the thickness of these layers and their survival time in terms of the rotation rate and stratification. Over longer timescales, layers merge to produce statistically steady states consisting either of a fully convective shell or of a convective layer ovelain by a persistent stably stratified layer (SSL), depending on the balance between stratification and rotational constraint. This shows that a layer subject to a semi-convection instability can self-organise into a convective layer and a SSL. These layers can generate a self-sustained magnetic field. Our dynamo simulations show that magnetic fields generated within the turbulent convective region are filtered by zonal flows in the overlying SSL, resulting in strongly a dipolar and axisymmetric external field, in encouraging agreement with Saturn's magnetic field. Across the explored parameter range, both the Rossby number and the thickness of the stably stratified layer are governed by a single combination of control parameters. This enable identification of a regime favourable to planetary-like dynamos.
arXiv:2605.26243v1 Announce Type: new
Abstract: Graph neural networks (GNNs) achieve strong performance on relational data, but real-world graphs are often distributed across organizations that cannot share raw data due to privacy and policy constraints. Existing federated GNN methods either ignore cross-client links, leading to degraded accuracy, or require frequent embedding exchanges, incurring substantial communication and privacy costs. We propose CE-FedGNN, a communication-efficient and privacy-preserving federated GNN framework for learning over such coupled graphs. Our approach avoids sharing raw data or per-round embeddings by infrequently exchanging aggregated node representations. To handle cross-client dependency and staleness, we introduce a moving-average estimator that continuously tracks node representations and enables their stable reuse across rounds. To provide formal privacy guarantees for the released representations, we adopt the metric differential privacy (metric-DP) framework, which measures privacy with respect to distances in the learned embedding space rather than worst-case input perturbations. This yields meaningful guarantees at noise levels where standard differential privacy becomes overly conservative. We establish convergence to a stationary point at a rate of $O(1/\sqrt{T})$ with $O(T^{3/4})$ communication complexity. In addition, we derive $(\varepsilon,\delta)$-metric-DP guarantees via R\'enyi differential privacy composition under a public-cohort threat model. Experiments on synthetic interbank anti-money laundering benchmarks and citation networks demonstrate that CE-FedGNN achieves strong performance while significantly reducing communication and maintaining robustness under privacy-preserving noise.
arXiv:2605.26256v1 Announce Type: new
Abstract: Multimodal large language model (MLLM)-based embodied agents have shown strong potential for solving complex tasks in physical environments. However, personalized assistance requires more than following generic instruction or recognizing object categories. In real-world scenarios, the intended target is often specified only implicitly through prior interactions, requiring agents to leverage personalized context accumulated over time. In this work, we propose POLAR, a multiomodal memory-augmented framework for personalized embodied agents over long-term user interactions. POLAR organizes prior interactions into a multimodal knowledge graph that captures semantic memory for personalized context and visual concepts, and episodic memory for embodied experiences such as agent trajectories. To execute embodied tasks, POLAR retrieves relevant memories to interpret the current request and guide task execution. We evaluate POLAR across multiple MLLM backbones and diverse evaluation scenarios to study the role of memory in long-term personalization. Results show that the proposed memory mechanism consistently improves performance by enabling more effective use of information accumulated over prior interactions. The gains are especially pronounced when the agents are required to reason across multiple interactions, perform multi-hop inference, or tracking updates in user-specific context over time.
arXiv:2601.08267v3 Announce Type: replace
Abstract: While reasoning-enhanced large language models perform strongly on English medical tasks, a persistent multilingual gap remains, with substantially weaker reasoning in local languages, limiting equitable global medical deployment. To bridge this gap, we introduce Med-CoReasoner, a language-informed co-reasoning framework that elicits parallel English and local-language reasoning, abstracts them into structured concepts, and integrates local clinical knowledge into an English logical scaffold via concept-level alignment and retrieval. This design combines the structural robustness of English reasoning with the practice-grounded expertise encoded in local languages. To evaluate multilingual medical reasoning beyond multiple-choice settings, we construct MultiMed-X, a benchmark covering seven languages with expert-annotated long-form question answering and natural language inference tasks, comprising 350 instances per language. Experiments across three benchmarks show that Med-CoReasoner improves multilingual reasoning performance by an average of 5%, with particularly substantial gains in low-resource languages. Moreover, model distillation and expert evaluation analysis further confirm that Med-CoReasoner produces clinically sound and culturally grounded reasoning traces.
arXiv:2601.08375v2 Announce Type: replace
Abstract: Semantic segmentation of 3D geospatial point clouds is fundamental to remote sensing applications, yet domain shifts caused by regional and acquisition-related variations often degrade model performance. Although domain adaptation can mitigate such shifts, existing methods typically require access to source-domain data, which is often infeasible due to privacy concerns and regulatory policies. To address this, we propose LoGo (Local-Global Dual-Consensus), a novel source-free unsupervised domain adaptation (SFUDA) framework requiring only a pretrained model and unlabeled target data. At the local level, we introduce a class-balanced prototype estimation module that ensures that robust feature prototypes can be generated even for sample-scarce tail classes, effectively mitigating the feature collapse caused by long-tailed distributions. At the global level, we introduce an optimal transport-based global distribution alignment module that formulates pseudo-label assignment as a global optimization problem, effectively correcting the over-dominance of head classes inherent in local greedy assignments, and thereby preventing model predictions from being severely biased towards majority classes. Finally, we propose a dual-consistency pseudo-label filtering mechanism that retains only high-confidence pseudo-labels where local multi-augmented ensemble predictions align with global optimal transport assignments for self-training. Extensive experiments on two challenging benchmarks, encompassing cross-scene and cross-sensor settings, demonstrate that LoGo consistently outperforms existing state-of-the-art methods. The source code is available at https://github.com/GYproject/LoGo-SFUDA.