arXiv:2605.28049v1 Announce Type: cross
Abstract: Designing compact and accurate circuits for the variational quantum eigensolver (VQE) is a central challenge in near-term quantum chemistry. Existing adaptive methods such as ADAPT-VQE design circuits by iteratively selecting operators from a predefined pool guided by gradient information and greedy heuristics. In this work, we adopt differentiable quantum architecture search (DQAS) as a circuit design framework based on the UCCSD operator pool, and introduce two complementary strategies: a global mode that simultaneously optimizes all operator selections, and a layerwise mode that constructs circuits incrementally while preserving previously learned structure. By relaxing discrete operator selection into a continuous differentiable optimization, DQAS enables gradient-based exploration over the combinatorial space of UCC circuit architectures. Benchmarks on BeH2, H4, LiH, H6, and H2O (8-14 qubits) show that both strategies achieve higher accuracy and fewer CNOT gates than ADAPT-VQE in the compact circuit regime, with up to 2.7-fold accuracy improvement for H2O and CNOT reductions of 13-17% at equivalent circuit depths. Benchmarks on the qubit-excitation-based (QEB) operator pool confirm that both advantages generalize beyond UCCSD. These results demonstrate that differentiable architecture search provides an effective and generalizable framework for designing accurate and compact VQE circuits in near-term quantum chemistry.
Science Journals
arXiv:2605.28313v1 Announce Type: new
Abstract: Large Language Models (LLMs) have demonstrated remarkable capabilities in tasks related to reasoning and judgment. However, assessing the quality of arguments requires a rigorous evaluation. We investigate the extent to which LLMs can effectively perform this task. We tested 12 open-weight LLMs of different sizes and families under zero-shot, few-shot, and chain-of-thought to approximate expert pairwise comparisons of argument quality across three dimensions-logical, rhetorical, and dialectic-and used these comparisons in a Bradley-Terry model to infer latent strength scores and derive a ranking of arguments. Our insights show that LLMs have promising but moderate correlation with human expert judgments, with Llama-70B obtaining the strongest alignment, reaching moderate Cohen's $\kappa$ = 0.493 and moderate correlations with Bradley-Terry scores derived from these annotations (Kendall, Pearson, and Spearman: 0.327-0.477). Other LLMs exhibit weak, moderate, or high alignment with Llama-70B while achieving comparable results against human experts, suggesting partial but complementary understanding of underlying quality dimensions despite differences in model size and family. Moreover, LLM predictions are stable across trial runs, with fewer than 7.75\% of cases yielding different labels. Remaining variability is handled via majority voting and few-shot prompting for large-size models.
arXiv:2605.28312v1 Announce Type: new
Abstract: Event-based vision sensors offer asynchronous, high-temporal-resolution measurements that are attractive for low-latency robotic perception, but many event-based motion estimation methods are computationally intensive and difficult to map to FPGA hardware. We present a streaming velocity estimator that discretizes asynchronous events into fixed-duration time bins, constructs a 1-bit spatial occupancy grid, and evaluates multiple velocity hypotheses in parallel using only fixed-width integer logic - shift registers, counters, comparators, and small LUT-mapped multiplies - with no dividers and no DSP blocks. It requires no frame reconstruction, no floating-point arithmetic, and no iterative optimization. The method deliberately trades dense sub-pixel optical flow for a sparse, quantized velocity estimate at each active pixel, suited to low-latency tasks such as reactive obstacle avoidance on size-, weight-, and power-constrained platforms. On noisy synthetic data with known ground-truth velocities, the method recovers both magnitude and direction, with magnitude estimates being most challenged when objects of different velocities intersect. On a real event-camera sequence, directional accuracy reaches 99.5% across all four evaluated motion segments, with performance remaining robust across occupancy densities in the 10-40% range. We characterize the algorithm's density-dependent behavior, present a parameter sensitivity analysis, show that the proposed datapath requires less than 2 kB of storage, and implement a single-axis prototype on a low-cost Xilinx Artix-7.
arXiv:2605.28309v1 Announce Type: new
Abstract: In large-scale benchmarking of stochastic optimization algorithms, the key challenge is no longer whether repeated runs are needed for reliability, but how to determine when sufficient evidence has been collected without incurring unnecessary computational cost. We study a learning-based extension of a recent empirical online heuristic that adaptively estimates the required number of runs using outlier handling and skewness-based symmetry checks. Using annotated outcomes from 132{,}000 Nevergrad runs on COCO (24 problems in 20 dimensions, 10 instances each, 11 optimizers), we train classifiers on 23 statistical, energy-free, and shape and stability features to predict whether a run-number estimate is reliable, prioritizing detection of incorrect estimates via minority-class recall. We evaluate reliability prediction using a within-configuration learning setup, where models are trained and tested on data sharing the same optimizer. The results show that run-number reliability can be learned in a within-configuration scenario, enabling detection of unreliable estimates with high minority-class recall, although performance remains limited by the restricted data diversity within fixed configurations.
arXiv:2604.04074v3 Announce Type: replace
Abstract: LLM-based reviewing systems typically take only the manuscript as input, leaving literature and code-based claims hard to verify. We present FactReview, a system that extracts review-relevant claims, grounds them in related work, and, when code is available, executes released artifacts under a fixed repair budget to audit empirical claims. Across 35 ML papers and 463 benchmark major claims, FactReview covers 84% of claims. Under an evidence-aware rubric, its reviews score 4.86/5 in overall quality, 0.7 above DeepReview-v2 and 1.5 above matched OpenReview comments. Removing execution evidence changes 17% of claim statuses, more than any other single evidence source. In a reviewer-assistance study, FactReview reduces mean review time by 58% while raising benchmark claim coverage from 87% to 99%. We argue that LLM reviewers should audit empirical claims, not make accept-reject decisions. The code is public at: https://github.com/DEFENSE-SEU/FactReview.
arXiv:2604.02028v2 Announce Type: replace
Abstract: Diffusion models have become a standard approach for generative modeling in continuous domains, yet their application to discrete data remains challenging. We investigate why Gaussian diffusion models with the DDPM solver struggle to sample from discrete distributions that are represented as a mixture of delta-distributions in the continuous space. Using a toy Random Hierarchy Model, we identify a critical sampling interval in which the density of noisified data becomes multimodal. In this regime, DDPM occasionally enters low-density regions between modes producing out-of-distribution inputs for the model and degrading sample quality. We show that existing heuristics, including self-conditioning and a solver we term q-sampling, help alleviate this issue. Furthermore, we demonstrate that combining self-conditioning with switching from DDPM to q-sampling within the critical interval improves generation quality on real data. We validate these findings across conditional and unconditional tasks in multiple domains, including text, programming code, and proteins.
arXiv:2510.06970v2 Announce Type: replace
Abstract: Compliance with maritime traffic rules is essential for the safe operation of autonomous vessels, yet training reinforcement learning (RL) agents to adhere to them is challenging. The behavior of RL agents is shaped by the training scenarios they encounter, but creating scenarios that capture the complexity of maritime navigation is non-trivial, and real-world data alone is insufficient. To address this, we propose a falsification-driven RL approach that generates adversarial training scenarios in which the vessel under test violates maritime traffic rules, which are expressed as signal temporal logic specifications. Our experiments on open-sea navigation with two vessels demonstrate that the proposed approach provides more relevant training scenarios and achieves more consistent rule compliance.
arXiv:2605.26189v2 Announce Type: replace
Abstract: Quantization-aware training (QAT) with low-bit floating-point formats enables efficient LLM deployment, yet introduces subtle failure modes invisible to standard training metrics. We present a systematic study of HiF8 W8A8 QAT for OpenPangu-Embedded-1B through the lens of Delayed Tensor Scaling (DTS). Across eight controlled experiments, we identify and disentangle two orthogonal failure modes: (i)amax saturation, where delayed scale estimates silently corrupt knowledge-sensitive representations via forward-pass clipping, and (ii)catastrophic forgetting, where an aggressive learning rate overwrites pretrained commonsense knowledge independently of quantization. Neither is detectable from training loss alone. We address amax saturation with a conservative max-algorithm DTS strategy over a 64-step history window, and mitigate forgetting via a 500-step BF16 warmup followed by QAT at lr=10^{-5}. Both fixes are necessary and sufficient: our final configuration achieves 0.43% MMLU drop, 0.58% HellaSwag drop, and 0.22% ARC-Challenge drop versus a matched BF16 baseline, with a training loss APE of only 0.11% over 10,000 steps.
arXiv:2605.26126v2 Announce Type: replace
Abstract: We generalize Tweed's seminal Born~I formalism for the electron-impact double ionization of two-electron atomic systems ($\text{H}^-$, $\text{He}$, $\text{Li}^+$) to non-extensive plasma environments described by Tsallis $q$-statistics. The bare Coulomb interaction is replaced by a $q$-deformed screened potential, which introduces a unique form factor envelope into the momentum-transfer integrations. We rederive the differential, partial, and total cross sections, demonstrating that this screening regularizes the classical high-energy Bethe logarithm and, in the $q < 1$ regime, induces diffraction-like oscillations in the angular cross sections. The model is calibrated against digitized historical experimental data to quantify how non-extensive screening deforms the vacuum baseline. Finally, coupling these cross sections to a $q$-Maxwellian distribution yields the macroscopic rate coefficients and reveals a novel subextensive critical-temperature threshold below which ionization is energetically forbidden.
arXiv:2510.02329v2 Announce Type: replace
Abstract: Speculative decoding accelerates LLM inference by verifying candidate tokens from a draft model against a larger target model. Recent judge decoding boosts this process by relaxing verification criteria by accepting draft tokens that may exhibit minor discrepancies from target model output, but existing methods are restricted by their reliance on human annotations or tasks with verifiable ground truths, limiting generalizability across diverse NLP tasks. We propose SelfJudge, which trains judge verifiers via self-supervision of the target model. Our method measures semantic preservation by assessing whether token-substituted responses preserve the meaning of original responses, enabling automatic verifier training across diverse NLP tasks. Our experiments show SelfJudge achieves superior inference-accuracy trade-offs than judge decoding baselines, offering a broadly applicable solution for faster LLM inference.
arXiv:2605.19444v2 Announce Type: replace
Abstract: Test-time reinforcement learning (TTRL) reports substantial accuracy gains on mathematical reasoning benchmarks using majority vote as a pseudo-label signal. We argue these gains are systematically misinterpreted: most reflect sharpening of already-solvable problems rather than genuine learning, while problems corrupted from correct to incorrect outnumber truly learned ones, and this damage is irreversible once majority vote locks onto a wrong answer. Per-problem tracking reveals that correct-answer signals in low-ability problems are briefly active before being permanently suppressed, a phenomenon we term the \textit{Correct-Answer Extinction Window}, with Flip Rate (FR) as its leading indicator. We thus propose TTRL-Guard, a lightweight framework with three mechanisms targeting the extinction window: Flip-Rate-Aware Reward Scaling (FRS) down-weights at-risk updates as FR declines, Minority-Preserving Sampling (MPS) retains gradient signal from minority correct answers, and Risk-Conditioned Sparse Updatings (RCSU) suspends updates on polarized problems. Experiments across three models and four benchmarks show that TTRL-Guard achieves the best average pass@1 on Qwen2.5-7B-Instruct and Qwen3-4B, improves relatively over TTRL by +54\% on AIME 2025.
arXiv:2605.11154v2 Announce Type: replace-cross
Abstract: Modern astrophysical studies rely heavily on complex data analysis pipelines; however, published descriptions often lack the detail required for computational reproducibility. In this work, we present an information-theoretic framework to quantify how effectively a method can be reconstructed from its written description. By treating algorithmic reconstruction as a probability distribution generated by Large Language Models (LLMs), we utilize Shannon entropy and Jensen-Shannon divergence to measure how strongly text constrains the hypothesis space of valid implementations. We demonstrate this approach through a case study of Trans-Neptunian Object (TNO) spectral reconstruction from sparse photometry. By prompting frontier LLMs with varying levels of manuscript text (Title, Abstract, and Methods), we find that while increasing text successfully clarifies the overall algorithmic structure, it fails to eliminate variance at the implementation level. This persistent variance establishes an "entropy floor," demonstrating that multiple divergent implementations remain consistent with explicit instructions. To evaluate practical reproducibility, we convert these reconstructed algorithms into executable pipelines. Our results reveal that, while LLMs easily recover core functional methodologies, they systematically fail to infer the tacit expert knowledge required for strict scientific calibration. This pilot study demonstrates that LLMs can be repurposed as a zero-shot diagnostic tool to audit methodological transparency, helping authors identify missing structural constraints and preserve scientific integrity in an era of automated research.
arXiv:2605.28478v1 Announce Type: new
Abstract: The commissioning of industrial electric drives still relies heavily on manual tuning of cascaded control loops, requiring expert knowledge and significant time. In this paper, we propose a fully automated approach for tuning the current control loop of industrial drives using Bayesian Optimization (BO) directly on real hardware, without requiring a system model or firmware modifications. The drive is treated as a black-box system, and the controller parameters are iteratively updated through closed-loop experiments. The tuning problem is formulated as a multi-objective optimization task that directly minimizes tracking error, time-weighted error, overshoot, and oscillatory behavior, enabling the identification of Pareto-optimal controller configurations. To address discrete parameters, noisy evaluations, and limited budgets, we adopt a multivariate Tree-structured Parzen Estimator (TPE) as the underlying BO strategy. The proposed method operates under practical industrial constraints, including communication latency and limited evaluation budgets. The experimental validation on a real motor drive system under no-load conditions shows that the method achieves performance comparable to expert tuning within a few minutes and without human intervention. Results show that Gaussian Process (GP)-based BO can yield highly competitive final solutions, but TPE-based BO is better aligned with this setting due to faster convergence, richer Pareto-front approximation, and lower computational overhead.
arXiv:2605.28468v1 Announce Type: new
Abstract: We present a hybrid robotic skin that combines electrical impedance tomography (EIT) with pneumatic tactile sensing to improve force reconstruction capability. The developed robotic skin is fabricated entirely by 3D printing and spray coating, making it affordable and easy to build. A Tikhonov-regularized inverse reconstruction, paired with per-pad pneumatic calibration, enables accurate large-area tactile sensing with a simple measurement scheme. For validation, we conducted load-cell indentation experiments; the results showed consistent force reconstruction across locations within a pad. Compared with an EIT-only baseline, sensitivity non-uniformity was also reduced, with the coefficient of variation decreasing from 0.31 to 0.14, indicating that the proposed approach addresses a longstanding limitation of EIT. We further demonstrated chest-mounted integration on a humanoid robot and found that the pneumatic signals remained reliable across diverse contact scenarios, including multiple simultaneous contacts on the same sensing pad. These results indicate a practical path toward accurate, scalable whole-body tactile sensing in real robotic systems.
arXiv:2605.28397v1 Announce Type: new
Abstract: Predicting conversion from Mild Cognitive Impairment (MCI) to Alzheimer's Disease (AD) is critical for early intervention. Current deep learning paradigms predominantly rely on cross-sectional structural MRI, neglecting prognostic value in patient-specific anatomical trajectories. We introduce the Temporal Adaptive Fusion Network (TAF-Net), a hybrid CNN-Transformer architecture that models paired longitudinal 3D MRI scans. Central to TAF-Net is a Temporal Fusion Module governed by an Adaptive Temporal Gate, which learns patient-specific weightings to synthesize three spatiotemporal representations: explicit structural change, region-to-region temporal cross-attention, and bilateral feature concatenation. Evaluated on the Alzheimer's Disease Neuroimaging Initiative cohort for three-year MCI-to-AD conversion prediction, TAF-Net achieved the highest discriminative performance among all evaluated methods using only structural MRI, significantly outperforming the strongest baseline and approaching multimodal methods requiring PET, CSF, or genetic data. The architecture exhibited exceptional data efficiency, matching baseline performance with a fraction of training data. Ablation studies demonstrate that longitudinal fusion improves discrimination while reducing predictive variance by 48% compared to single-timepoint evaluation. Interpretability analyses reveal spatial attention aligned with established AD pathology in the medial temporal lobe and ventricles, while the gating mechanism prioritizes explicit volumetric change with strong positive correlation to conversion risk.
arXiv:2605.28571v1 Announce Type: new
Abstract: Despite warnings that LLMs can make mistakes, users often develop inappropriate trust and accept incorrect answers without critical evaluation. Uncertainty quantification (UQ), displaying LLMs' confidence, has emerged as a promising approach to calibrate user trust. However, prior empirical studies on uncertainty communication have treated uncertainty as a single numerical score or simple natural language expression. This simplification fails to capture a key property of LLM outputs: a single response often comprises multiple claims and reasoning steps, each with distinct levels of uncertainty. To address this gap, this study investigates uncertainty granularity (i.e., the extent to which uncertainty is expressed at different levels within an LLM response) and examines its impact on LLM-assisted decision-making. We conducted a large-scale, between-subjects study (N=192) in which participants answered medical questions using LLMs that displayed uncertainty at three different granularities: output-level (entire response), relation-level (individual reasoning steps), and token-level (specific words). Our findings reveal distinct behavioral effects as a function of uncertainty granularity. Token-level uncertainty increased users' agreement with the AI, whereas output- and relation-level uncertainty did not increase agreement but instead reduced users' confidence in their own answers. Notably, relation-level uncertainty also reduced external verification (i.e., internet searches, checking provided URLs), steering users away from independent fact-checking and toward reliance on the LLM and its accompanying uncertainty cues. Our findings demonstrate that uncertainty granularity significantly shapes how users interact with and verify LLM outputs, providing concrete design guidance for building responsible LLM applications that encourage appropriate skepticism and verification behaviors.
arXiv:2504.09637v3 Announce Type: replace
Abstract: We establish optimal convergence rates for the continuous piecewise affine finite element approximation of the Sobolev constant in arbitrary dimensions N\geq 2 and for Lebesgue exponents 1<p<N. Our analysis relies on a refined study of the Sobolev deficit in suitable quasi-norms, which have been introduced and utilized in the context of finite element approximations of the p-Laplacian. The proof further involves sharp estimates for the finite element approximation of Sobolev minimizers.
arXiv:2512.09786v2 Announce Type: replace
Abstract: Examples of embedded intelligence include a wide variety of tiny neural networks used on-board wireless sensors and actuators, which are expected to continuously perform inference on time-series of the data they sense. In order to fit lifetime and energy consumption requirements when operating on battery, such hardware is exclusively based on microcontroller with as little memory as possible, e.g., 128 kB of RAM. In this context, optimizing data flows during inference across neural network layers becomes crucial. In this paper, we introduce a new framework, TinyD\'ej\`aVu, and novel algorithms we designed to drastically reduce the RAM budget required by inference using various neural network models for sensor data time-series on typical microcontroller hardware. We publish the implementation of TinyD\'ej\`aVu as open source, and we perform reproducible benchmarks on common microcontroller hardware (Arm Cortex-M). We show that TinyD\'ej\`aVu can save up to 90\% of RAM usage with equal compute latency compared to prior work (StreamiNNC) on overlapping sliding window inputs.
arXiv:2512.20657v2 Announce Type: replace
Abstract: The source detection problem arises when an epidemic process unfolds over a contact network, and the objective is to identify its point of origin, i.e., the source node. Research on this problem began with the seminal work of Shah and Zaman in 2010, who formally defined it and introduced the notion of rumor centrality. With the emergence of Graph Neural Networks (GNNs), several studies have proposed GNN-based approaches to source detection. However, there is room to strengthen methodological clarity and reproducibility across these works. As a result, it remains unclear whether GNNs truly outperform more traditional source detection methods across comparable settings. In this paper, we first systematically review existing GNN-based methods for source detection, clearly outlining the specific settings each addresses and the architectures they employ. We then reproduce and benchmark four representative GNN architectures against a diverse set of traditional and MLP-based baselines under controlled, comparable conditions. We also investigate key questions surrounding this problem, including how detectability evolves over time, how performance scales with training set size, and how sensitive methods are to uncertainty in observation timing and epidemic parameters. Our experiments show that GNNs substantially outperform all other methods we test across a variety of network topologies. Although we initially set out to challenge the notion of GNNs as a solution to source detection, our results instead demonstrate their remarkable effectiveness for this task. To ensure full reproducibility, we release all code and data on GitHub. Finally, we argue that epidemic source detection constitutes a natural and attractive benchmark task for evaluating GNN architectures.
arXiv:2605.12317v2 Announce Type: replace
Abstract: Popular centroid-based clustering methods are typically optimized for global objectives, and may fail to adequately represent large groups of datapoints. Thus, one needs proportionality notions suited for metric settings. Ideally, such notions should admit polynomial-time algorithms for (a) finding proportional outcomes, and (b) checking if a given outcome is proportional; the latter enables evaluation of traditional algorithms without proportionality guarantees (e.g., $k$-means). A promising approach imports proportionality notions from multiwinner voting with approval ballots. In particular, mPJR, the metric version of the well-known Proportional Justified Representation (PJR) axiom, satisfies (a), but whether it satisfies (b) was open.
In this work, we study the computational complexity of auditing proportional representation in clustering. In the approval setting, PJR is coNP-complete to verify; however, it admits a strengthening PJR+, which satisfies (a) and (b). We show these results translate to the metric setting: mPJR is coNP-complete to verify, we define mPJR+, a metric analog of PJR+, and argue mPJR+ satisfies (a) and (b). However, auditing mPJR+ relies on repeated submodular minimization, rendering it impractical at scale, and a natural combinatorial approach is infeasible.
As a partial remedy, we propose an mPJR+ verification algorithm exponential in $k$ but quasilinear in the number of datapoints. Motivated by these hardness results, we introduce DC-mPJR+: a proportionality concept offering representation guarantees to a restricted set of coalitions around unselected centers, admitting an $O(mn \log n + mnk)$ verification algorithm. DC-mPJR+ outcomes can be computed efficiently, and any $\gamma$-DC-mPJR+ solution satisfies $(\gamma + 2)$-mPJR+.
Sign-Aware Gated Sparse Autoencoders: Modeling Anticorrelated Features with Bi-Jump-ReLU Activations
arXiv:2605.28149v1 Announce Type: new
Abstract: Sparse Autoencoders (SAEs) extract interpretable features from Large Language Models, but standard variants enforce non-negativity, forcing separate latents for diametrically opposed concepts (e.g., "pressure too high" vs. "pressure too low") and wasting dictionary capacity when features are anticorrelated. We propose the Sign-Aware Gated SAE (SA-GSAE): two-sided gated sparsity with signed magnitude and auxiliary supervision. A polarity-sensitive gate selects support on either sign, a signed-magnitude path avoids L1 shrinkage, and an auxiliary reconstruction prevents gate collapse. Bipolar sharing - one latent encoding both signs along a shared direction - is realised via a new Bi-Jump-ReLU activation; parameter accounting shows sign-awareness stays parameter-efficient even when anticorrelated pairs are rare.
On real LLM activations across three mid-depth hookpoints on Pythia-1B and SmolLM3-3B (6 cells, 3 seeds), a half-width SA-GSAE at width H strictly Pareto-dominates a full-width Gated SAE at 2H over the entire swept L0 overlap on 3 of 6 cells (both MLP-output hookpoints and resid-mid/Pythia-1B); on the remaining 3 it matches R^2 within 0.025 (max gap -0.008) while cutting dead fraction by 0.35-0.62 absolute. Sweep-geomean dead-fraction reductions are ~100x-500x on MLP-output cells and Pythia-1B resid, ~2x-4x on attention cells and SmolLM3-3B resid. Ablations show the two-sided gate and auxiliary loss are load-bearing (no auxiliary collapses LR to 0.27, 98% dead); tying r_i^+ = r_i^- is indistinguishable (|Delta R^2| = 0.0015), and we recommend this symmetric variant as default. MLP-output gains come from most latents carrying both polarities; on attention, bipolar structure concentrates in a small set of top latents. Full-width SA-GSAE exhibits a reproducible reconstruction collapse at SmolLM3-3B resid that the half-width entirely avoids.
arXiv:2605.28115v1 Announce Type: new
Abstract: Vision-Language Models (VLMs) face severe memory and latency bottlenecks due to high-resolution visual tokens. While current token reduction methods theoretically save FLOPs, post-hoc pruning introduces structural overhead, failing to yield proportional wall-clock acceleration. However, enforcing a contiguous compact pathway risks geometric disorientation and loss of fine-grained localization. To overcome these barriers, this paper introduces CIVIC, a path-consistent compact visual inference framework. By maintaining compact sequence representations seamlessly across the vision encoder, projection layer, LLM prefill, and KV-cache, CIVIC avoids non-contiguous memory access and localized unmerging overheads. Evaluated on the Qwen3-VL architecture, CIVIC successfully translates sequence reductions into genuine physical hardware efficiency, shrinking KV-cache memory to approximately one-third of the baseline and reducing end-to-end inference latency. Enabled by text-aligned KL distillation and an adaptive spatial retention floor, CIVIC achieves these efficiency milestones without degrading accuracy across rigorous multimodal reasoning and visual grounding benchmarks.
arXiv:2605.28112v1 Announce Type: new
Abstract: Federated Retrieval-Augmented Generation (FedRAG) is attractive for privacy-sensitive applications because raw data remain local. As a result, routing must rely on client-provided semantic profiles, creating a new opportunity for manipulation. We introduce Routing Hijacking, a routing-stage attack in which a malicious client forges its profile to attract target queries despite having irrelevant underlying data. We show that this vulnerability is severe. Across three representative FedRAG routing architectures, Routing Hijacking consistently misroutes target queries and leads to downstream disruptions and failures, including missing evidence, poisoning, incorrect answers, and hallucinations. In a high-stakes MedQA-USMLE case study, we further show that poisoned retrieved evidence can mislead models across scales, leading to incorrect answers, hallucinations, and sycophantic failures. Existing defenses do not close this gap: encrypted routing preserves the exploited ranking, and Byzantine-robust Federated Learning (FL) rules transfer poorly to heterogeneous routing profiles. To address this gap, we propose a trust-aware post-routing framework that reweights clients using returned-evidence feedback, including retrieval relevance, profile consistency, and cross-client agreement; online experiments show that it suppresses persistent hijacking over recurring queries and transfers to a learned neural router. Our findings establish routing integrity as a new security challenge in FedRAG and highlight the need for stronger defenses for secure federated retrieval.
arXiv:2605.28110v1 Announce Type: new
Abstract: With the rapid development of simulation tools, the development and validation of autonomous robotic systems have become more efficient before real-world deployment. This paper presents a simulation-to-real implementation of an autonomous mobile robot based on an existing mechanical platform. Instead of focusing on mechanical design, our work concentrates on the development of the onboard control, self-localization, and autonomous navigation system. The proposed robot is equipped with onboard sensing and computation to estimate its pose and navigate autonomously in the environment. The overall framework is first developed and tested in simulation, and then deployed on the real robot for experimental evaluation. The results demonstrate the feasibility of the proposed approach and show that simulation provides an effective foundation for developing reliable autonomous mobile robot systems. The source code will be released at https://ntdathp.github.io/outdoor-robot-web.
arXiv:2605.28096v1 Announce Type: new
Abstract: Primary hemostasis is initiated by platelet adhesion and aggregation at a site of vascular injury and is strongly regulated by local hydrodynamic conditions. At elevated shear rates, platelet capture is mediated by von Willebrand factor (vWF), a multimeric protein that undergoes shear-induced unfolding and becomes adhesive. We investigate early-stage clot formation under physiological high-shear-flow conditions by employing particle-based mesoscale hydrodynamics simulations with explicitly resolved red blood cells, platelets, and mechano-sensitive vWF in a microchannel geometry. The model incorporates vWF-mediated adhesion of platelets to a hemostatic surface, together with non-periodic inflow-outflow boundary conditions that allow continuous material supply and transport. We analyze the dynamics of platelet-vWF aggregation, clot growth dynamics, clot geometry and internal stresses, and thrombo-embolization across a range of elevated flow rates. Our results demonstrate that clot formation proceeds through the establishment of platelet-vWF aggregates at the hemostatic site, and that the clot reaches a finite size determined solely by hydrodynamic forces, without invoking biochemical stabilization mechanisms. Beyond a critical size, increased drag from fluid flow leads to recurrent embolization events that limit further growth. These findings highlight the central role of hydrodynamic stresses in regulating primary hemostasis and provide a mechanistic framework for understanding clot stability under physiological flow conditions.