Peer Revenue Publish or profit

Papers

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

All of arXiv

Monday 7 September

Analysis of PDEs

Asymptotic stability of viscous shock for the outflow problem of 3D Navier-Stokes system

We establish the asymptotic stability of planar viscous shock waves for the outflow problem of the three-dimensional barotropic compressible Navier–Stokes equations in a half-space, with periodic boundary conditions imposed in the transverse directions. For a weak shock located sufficiently far from the boundary, we prove that, under small perturbations in $H^{2}$, the outflow problem admits a unique global-in-time solution and that the solution converges uniformly to the viscous shock, up to a dynamical shift, whose velocity time-asymptotically decays. This provides the first shock-stability result for the multidimensional outflow problem. Our proof combines the $a$-contraction method and higher-order energy estimates adapted to the half-space boundary. The boundary trace remaining in the zeroth-order estimate is controlled by differentiating the perturbation system in the tangential directions and exploiting the favorable sign of the outflow flux. The highest-order normal derivatives are then recovered from the time and tangential derivatives through the momentum equation, while the boundary terms generated by the dynamical shift are controlled using the exponential decay of the viscous shock tail.
Read the full abstract

Computer Vision

Continuous Token-Level Spatio-Temporal Context Modeling for Visual Object Tracking

Spatio-temporal context has become increasingly crucial for visual tracking. However, most existing approaches extract spatio-temporal cues via discrete sampling strategies, which inherently deviate from the continuity of spatio-temporal context, thereby deteriorating tracking performance. To address this challenge, we propose TLCTrack, a novel tracking framework that models token-level spatio-temporal context through continuously updated salient tokens, enabling more accurate target representation. Specifically, TLCTrack incorporates three components: Masked Unidirectional Attention (MUA), Spatial Salient Token Collection (SSTC), and Temporal Salient Token Bank (TSTB) modules. By explicitly integrating spatio-temporal context, MUA extracts discriminative targetaware spatial features in the search region. To avoid the negative impact of background on feature learning, SSTC progressively suppresses background interference, thereby enhancing target spatial representation. Finally, TSTB captures high-quality spatio-temporal information through continuous salient token updates. Extensive experiments on five benchmarks demonstrate that our method achieves superior performance over state-of-the-art trackers. Code and models are available at https://github.com/xiading123/TLCTrack.
Read the full abstract

All authors Ding Xia, Meiqin Liu, Jing Zhou, Jian Lan

Algebraic Geometry

A note on compact Kähler varieties with pseudo-effective tangent sheaf

In this paper, we discuss recent developments and open problems concerning the geometry of compact Kähler varieties whose tangent sheaves are pseudo-effective in a strong sense. As our main result, we prove that, after passing to a finite quasi-étale cover, any compact Kähler variety with quotient singularities and pseudo-effective tangent sheaf admits a flat fibration onto a complex torus, with rationally connected fibers. We also obtain an analogous structure theorem for klt compact Kähler varieties, assuming the existence of minimal models in numerical dimension zero; consequently, the result holds unconditionally in the projective setting and for low-dimensional compact Kähler varieties.
Read the full abstract

High Energy Physics (Phenomenology)

Exploring the $P$-wave bottom-strange molecular pentaquarks with the complex scaling method

In this work, we perform a systematic investigation of the $P$-wave $\Sigma_b K/\Lambda_b K^{*}/\Sigma_b K^{*}$ and $\Sigma_b\bar{K}/\Lambda_b\bar{K}^{*}/\Sigma_b\bar{K}^{*}$ systems for all possible $I(J^P)$ combinations. In contrast to our earlier study of the negative-parity sector [Eur. Phys. J. C 85, 1026 (2025)], the positive-parity sector is dominated by resonances. For the coupled-channel $I(J^P)=1/2(1/2^+)$ $\Sigma_b K/\Lambda_b K^*/\Sigma_b K^*$ system, we predict a narrow resonance above the $\Sigma_b K$ threshold. In the $I(J^P)=1/2(3/2^+)$ $\Lambda_b K^*/\Sigma_b K^*$ system, a molecular resonance candidate is found above the $\Lambda_b K^*$ threshold. For the higher-isospin $I=3/2$ sector, resonances are identified in the $\Sigma_b K^*$ subsystem for both $J^P=1/2^+$ and $3/2^+$ cases, though both states exhibit relatively compact sizes. We also extend our analysis to the $\Sigma_b\bar{K}/\Lambda_b\bar{K}^{*}/\Sigma_b\bar{K}^{*}$ systems, where a cutoff-sensitive resonance is found in the $I(J^P)=1/2(1/2^+)$ configuration. A broad molecular resonance candidate is found below the $\Sigma_b\bar{K}^*$ threshold in the $I(J^P)=1/2(3/2^+)$ system. We hope that these predictions provide useful guidance for future experimental searches for bottom-strange molecular pentaquarks by LHCb.
Read the full abstract

Probability

Optimal central limit theorem for bounded random variables in high dimensions

Let $W=n^{-1/2}\sum_{i=1}^n X_i$, where the $X_i$ are independent centered random vectors in ${\mathbb R}^p$ with $|X_{ij}|\le B$ almost surely. Suppose that $\text{Cov}(W)$ has unit diagonal and smallest eigenvalue at least $b^2>0$. We prove that the distance between $W$ and a Gaussian vector with the same covariance, uniformly over axis-aligned rectangles, is at most $C\min\{1,b^{-2}Bn^{-1/2}\log^{3/2}(ep)\}$. For fixed $b$, the dependence on summand size and dimension matches known lower bounds in growing-dimensional regimes. The proof combines a concentration estimate near rectangle boundaries with a carefully chosen Gaussian comparison.
Read the full abstract

Probability

Optimal Sobolev Rate for Gaussian Density Approximation of Wiener Chaos Vectors

Let $(F_n)$ be a sequence of random vectors with identity covariance matrix whose components belong to the same fixed Wiener chaos, and assume that $F_n$ converges in law to a standard Gaussian vector. We prove that, for every integer $m\geq0$ and every $p\in[1,\infty]$, the optimal rate of convergence of the densities in the Sobolev space $W^{m,p}$ is given by the maximum of the absolute third-order cumulants and the diagonal fourth-order cumulants. The same quantity also gives the optimal rates in total variation, Kolmogorov and $1$-Wasserstein distances. Our proof first derives, by Gaussian interpolation and Gaussian convolution, a cumulant expansion in the space of tempered distributions without imposing Malliavin nondegeneracy at the endpoint. Finite-order Malliavin density estimates then upgrade this identity to Sobolev spaces. Matching lower bounds follow from a finite-dimensional argument in which parity separates the third-order and fourth-order Gaussian corrections, while norm equivalence rules out cancellations among mixed cumulants. A superconvergence theorem supplies the required finite negative moments of the Malliavin determinant along a sufficiently far tail of the approximating sequence, so that no Malliavin nondegeneracy assumption is required.
Read the full abstract

Artificial Intelligence

VST: Verifiable Structured Transport for Auditable Agent-to-Agent Alpha Discovery

Agent-to-agent (A2A) alpha discovery is slowed by repeated feedback cycles between mining and evaluation agents, whose hand-offs, in contemporary LLM multi-agent systems, are free-form natural-language messages that carry no stable contract and cannot be replayed. We first restructure this communication as a structured agent-to-agent protocol of typed, causally addressable, unicast records, so that the committed stream forms a causal trajectory. On that trajectory a single predictor with four typed heads forecasts the accumulated guidance the two miners would receive several cycles ahead; a transactional verify–leap controller then commits a multi-cycle speculative outcome only when it passes a four-level gate, and otherwise rolls back to the exact prior state. Structure is the enabling contribution, and its value is not accuracy. A controlled ablation shows an equal-information free-text channel reaches the same predictor hit rate. What typing provides is a state that can be schema-checked, replayed deterministically, and prevented by construction from leaking a forecast to an evaluator: auditability by construction, not an empirically stress-tested guarantee. On a CSI~1000 out-of-sample holdout, our single run is the only one among eight methods (seven baselines and ours) to hold a positive median annualized return and Sharpe at the factor level, though the median return in excess of the benchmark stays negative for every method including ours; its development-selected top-20 portfolios reach a $0.71$ median holdout Sharpe, selected on a split inside the optimization horizon. We report these single-run results descriptively, gross of costs, and are explicit about their limits throughout; in particular we do not isolate the effect of the leap machinery from the inherited search substrate, which we leave to future work.
Read the full abstract

Computer Vision

SpatialBlock: Enhancing Spatial Intelligence in LVLMs via Synthetic Block-Stacking Problem

Large Vision-Language Models (LVLMs) have achieved strong performance on diverse visual tasks, yet their ability to reconstruct and reason about the 3D structure of the scene depicted in 2D images – referred to as spatial intelligence – remains limited. Existing approaches attempt to address this gap by using real-scene spatial question answering datasets that require dense geometric annotations. However, constructing such labels is costly, time-consuming, and often noisy due to reliance on external perception modules. In this work, we propose a novel paradigm inspired by human cognitive development: learning foundational spatial skills through structured block-manipulation tasks. We introduce SpatialBlock-15k, a synthetic dataset of 15,000 block-stacking problems covering 3D-to-2D projection, viewpoint transformation, and structural combination. The dataset further incorporates controlled color modulation as visual cues to encourage anchor-based reasoning in visually complex conditions. Experiments demonstrate that LVLMs trained on our dataset through either direct answering or reasoning-based prediction significantly outperform baselines and generalize to real-world spatial tasks, despite the dataset's synthetic and compact nature. Code and data are available at https://github.com/rsoohyun/SpatialBlock.
Read the full abstract

Machine Learning

Trust-But-Verify: Poisoning-Resilient Locally Private Graph Learning Protocols

Built upon local differential privacy (LDP), locally private graph learning protocols have emerged as an important paradigm for decentralized graph learning, balancing privacy protection and learning utility. Under such protocols, each user locally perturbs their node features and adjacency information before transmission, ensuring formal privacy guarantees without original data leaving the device. However, the inherently open participation nature renders these protocols critically vulnerable to data poisoning attacks, where adversaries inject carefully crafted malicious nodes to corrupt neighborhood aggregation and degrade downstream utility. Despite the severity of this threat, effective defenses in this setting remain largely unexplored. In this paper, we propose VERITAS, a poisoning-resilient locally private graph learning protocol built on a trust-but-verify paradigm. By introducing a verification list encoding graded peer trust levels, VERITAS jointly privatizes node features and graph structure on the user side, while exploiting bilateral attestation asymmetry on the server side to identify and prune malicious nodes. Concretely, VERITAS comprises four synergistic stages: (1) local data perturbation, (2) attestation-driven malicious node pruning, (3) utility restoration via dual denoising, and (4) robust private graph learning. Extensive experiments on four real-world benchmark datasets across multiple LDP mechanisms and GNN architectures demonstrate that VERITAS effectively defends against data poisoning attacks and significantly improves downstream graph learning utility under rigorous privacy guarantees.
Read the full abstract

All authors Longzhu He, Li Sun, Hao Peng, Ruijie Wang, Raymond Chi-Wing Wong, Sen Su

Game Theory

How Well Can Strategyproof Tournament Rules Resist Pairwise Manipulation?

A tournament rule maps the outcomes of all pairwise matches among $n$ teams to a possibly randomized winner. Desirable rules should be Condorcet consistent and monotone, yet also resistant to manipulation among coalition. Prior work mostly measures such manipulation additively through $k$-strongly non-manipulable at $\alpha$ ($k$-SNM-$\alpha$), meaning that no coalition of size $k$ can fix the matches among themselves to increase their total winning probability by $\alpha$. Very recently, two new notions of non-manipulability were introduced. Multiplicative non-manipulability ($k$-MNM-$\delta$) is defined analogously, using the multiplicative factor instead. Non-manipulability for $\lambda$ ($k$-NM$_\lambda$) characterizes the selfishness of a team, which restricts a coalition's gain to be less than $\lambda$ times the winning probability sacrificed by its members. In this work, we begin with a strict hierarchy among these three notions: NM$_\lambda$ is stronger than MNM, which is then stronger than SNM. This motivates us to consider those two notions that are stronger but less studied: pairwise multiplicative non-manipulability and $2$-non-manipulability for $\lambda$. We show that Randomized Death Match is $2$-MNM-$3/2$ and optimally matches the lower bound. Then, we introduce the BlockBonusedWinStrengths rule, which is Condorcet consistent, monotone, and $2$-NM$_2$. This rule substantially improves the previous upper bound of $\lambda=11$ and comes within a factor of two of the lower bound $\lambda=1$.
Read the full abstract

Machine Learning

PhysSAE: Mechanistic Interpretability of PINNs with Sparse Autoencoders

Physics-Informed Neural Networks (PINNs) embed PDE residuals into neural network training, but their internal representations remain opaque: it is unknown what physical features their hidden layers encode or whether those features have a localized causal role. We present PhysSAE, a mechanistic interpretability framework that trains overcomplete sparse autoencoders (SAEs) on PINN penultimate-layer activations and evaluates dictionary atoms through direct causal intervention in the original frozen hidden state: $h_{\mathrm{cf}} = h - \alpha z_k d_k$, bypassing the SAE decoder entirely. Across six PDE families, with 3 PINN seeds and 3 SAE seeds each—we show that (i) Our discovered SAE atoms align with independently-defined physical observables (max Pearson $|r|=0.951$, always $\gg$ permutation null), (ii) the causal footprint of top-aligned atom ablation is 1.2–4.2$\times$ more spatially concentrated canonical than PCA or ICA interventions, and (iii) top-aligned atoms outperform matched random controls on causal localization for structured physical concepts (ESF$_{80}$ advantage 0.04-0.44). Two-atom bilateral representations improve concept regression R$^2$ by $\Delta R^2\!=\!0.05\text{-}0.15$ over single atoms, while random pairs decrease it by up to 0.60. These results demonstrate that PINNs develop sparse, physically structured latent representations that can be identified and causally interrogated post-hoc, opening a path toward interpretability-aware scientific machine learning.
Read the full abstract

Quantum Physics

2-Fold Forrelation is in QAC$^0$

We show that 2-fold Forrelation with inverse-polylogarithmic promise gap can be solved, with bounded error, by polynomial-size QAC$^0$ circuits. Unlike the standard oracle-based Forrelation algorithm, our circuits receive the input explicitly, in the same form as the AC$^0$ circuits against which Forrelation is known to be hard. At constant gap, this yields a natural promise-problem separation between QAC$^0$ and AC$^0$.
Read the full abstract

Hardware Architecture

NOVA-CIM: Noise- and Correlation-Tolerant Stochastic Interfaces for Analog Compute-in-Memory

Analog compute-in-memory (CIM) enables energy-efficient model acceleration, but its reliance on ADC-based readout, which directly quantizes noisy column currents, makes inference accuracy highly sensitive to analog read noise, active-row scaling, and ADC precision. In this paper, we present NOVA-CIM, a noise- and correlation-tolerant stochastic interface for analog CIM by replacing multi-bit ADC readout with random-reference 1-bit sensing and reconstructing results through lightweight counting. By converting column currents into comparison probabilities, this probability-domain readout averages zero-mean dynamic read noise over stochastic samples while reducing dependence on high-resolution ADCs. We provide a unified robustness analysis showing that dynamic read noise is suppressed through temporal averaging and that spatial input-bitstream correlation increases instantaneous current variance rather than introducing first-order MAC bias. MAC-level experiments and end-to-end evaluation on ViT-Base validate the analysis: under read noise, Top-1 accuracy remains 84.48% near the 84.51% bfloat16 (BF16) baseline; under stochastic number generator (SNG) reuse, MAC bias stays near zero while root-mean-square error (RMSE) and stochastic cross-correlation (SCC) grow as predicted.
Read the full abstract

All authors Jiachen Ren, Wenshuai Yao, Haobo Liu, Xincheng Feng, Chenxi Hu, Zhengwu Liu, Kechao Tang, Wenyong Zhou, Ngai Wong

Computer Vision

CNsEMD: An Expert-Annotated Multi-Field-Strength MRI Dataset and a Hyperspherical Manifold Network for Multimodal Cranial Nerve Parcellation

Cranial nerves (CNs) play essential roles in sensory, motor, and autonomic functions. Accurate CN parcellation from multimodal magnetic resonance imaging (MRI) is crucial for neuroanatomical analysis and neurosurgical planning. However, accurate CN parcellation remains extremely challenging because CNs are very small, exhibit low image contrast, and have slender tubular morphologies and complex anatomical trajectories. Moreover, the lack of publicly available, expert-annotated datasets has impeded the development and fair benchmarking of learning-based CN analysis methods. In this work, we introduce CNsEMD, an expert-annotated multimodal dataset for CN parcellation. It comprises data from 202 subjects acquired on 3T, 5T, and 7T MRI scanners. We further propose the projective hyperspherical manifold network (PHM-Net), which learns cross-modal representations by capturing angular relationships in a shared hyperspherical embedding space. Rather than performing multimodal fusion in Euclidean space, the proposed Hyperspherical cross-modal interaction (HCI) module enables bidirectional feature exchange between T1-weighted (T1w) and direction-encoded color (DEC) representations on a unit hypersphere. The Magnitude-preserving projective hyperspherical orientation representation (PHOR) captures the axial nature of DEC orientations while preserving diffusion magnitude. The hyperspherical prototype segmentation head (HPSH) further extends angular similarity to voxel-wise classification using normalized voxel embeddings and learnable class prototypes. Extensive experimental results on the CNsEMD dataset demonstrate the effectiveness of our PHM-Net against state-of-the-art methods. CNsEMD establishes a reproducible benchmark for multimodal CN imaging, while PHM-Net provides a geometry-consistent solution for CN parcellation across diverse MRI acquisitions.
Read the full abstract

All authors Lei Xie, Junxiong Huang, Guoqiang Xie, Jiawei Zhang, Jiahao Huang, Tianling Lyu, Ye Wu, Mingchu Li, Shoujun Yu, Shanshan Wang, Qingrun Zeng, Yuanjing Feng

Earth and Planetary Astrophysics

The First Observations of Moonlit Satellites

Optical Space Domain Awareness (SDA) operations have traditionally relied on Sunlit passes near the terminator, limiting continuous tracking of low-Earth orbit (LEO) satellites. In this work, we show that by accounting for all four major illumination sources: Sunlight, Earthshine, Moonlight, and Lunar-Earthshine, large LEO satellites like the ISS can be optically monitored for a full 24 hour period. We present what we believe to be the first quantitative observations of the International Space Station (ISS) and Chinese Space Station (CSS) illuminated only by Moonlight and Lunar-Earthshine under night-time conditions. Using a 0.6 m telescope, we obtain 147 detections of the ISS with a median brightness of V = $12.02 \pm 0.17$ mag, approximately $13.1 \pm 1.3$ magnitudes fainter than in daylight. For interpretation, we extend a satellite brightness model based on reflected Sun light (lumos-sat) to include Moonlight and Lunar-Earthshine, and validate this model both against the new observations and independent Ansys Systems Tool Kit simulations, achieving residuals of $0.03 \pm 0.80$ mag. The simulations confirm that Lunar-Earthshine dominates the illumination of nadir- facing components, boosting their radiance by at least 100 times relative to Starlight alone. Applying this validated model to existing and proposed large satellites, we show that Moonlight and Lunar-Earthshine will make these future spacecraft well within detection limits of even modest ($\sim 0.6$ m) optical SDA systems for most of the night, and for up to $\sim 11$ nights per lunar month at local midnight. When Moonlit observations are combined with daytime and twilight observations, this enables near-continuous ($\sim 24$ h) optical monitoring of LEO satellites from a single site, significantly increasing SDA tracking capabilities.
Read the full abstract

All authors Sarah E. Caddy, Stephen E. Catsamas, Michele Trenti, Lee R. Spitler, Patrick North

Systems and Control

Lightweight Onboard Routing and Topology Co-Construction for Autonomous GNSSs

Global navigation satellite systems (GNSSs) can already plan their inter-satellite link (ISL) topology onboard, leaving routing as the last step toward autonomous networking. In GNSS ISLs, each satellite links to at most one peer per time slot. We exploit this structure to replace end-to-end path search with a single backward sweep over the slots. The proposed method, DS-SMFP, propagates earliest-arrival times backward from each destination and outputs a STORE/FORWARD decision for every slot, satellite, and destination, in time linear in the number of slots and with working memory independent of the planning horizon. Because each slot needs only its own topology and the delivery state of the next one, routing can steer link selection while the topology is built. This slotwise dependency enables Master-Oriented Topology–Routing Co-Construction (MTRCC), which favors traffic to and from a master satellite performing onboard orbit determination. In a 24-h, 30-satellite BeiDou scenario, DS-SMFP matches the delays of Contact Graph Routing (CGR) and shortest path tree (SPT) routing while running over 180 times faster and using over 36 times less peak routing working memory. Compared with fair contact plan (FCP) topology, MTRCC reduces mean master-related traffic delays by over 10%, with an approximately 3% increase in mean PDOP. This paper provides a methodological foundation for efficient autonomous GNSS networking.
Read the full abstract

All authors Huan Yan, Juan A. Fraire, Rongqin Yang, Wenfeng Li, Kanglian Zhao

Mesoscale and Nanoscale Physics

Theory of staggered damping-like spin-orbit torque on synthetic antiferromagnetic domain walls

The Néel, or staggered, spin-orbit torque (SOT) enables an efficient electrical manipulation of antiferromagnetic order. However, it has been tied to special crystal structures, and only its field-like component is established. We develop a theory of staggered damping-like SOT, which arises generically in synthetic antiferromagnets sandwiched by the same heavy metal on both sides. Deriving a Néel-vector Lagrangian, we obtain the domain-wall (DW) velocity in closed form: bilinear in current and in-plane field, opposite in sign for Néel and Bloch walls, and tunable through the interlayer exchange coupling. DW structure thus becomes a control knob for DW motion, while the DW motion, in turn, offers an electrical readout of its structure. Micromagnetic simulations confirm the theory. The predicted sign reversals and scaling laws of the DW velocity, together with the thresholds for the Néel-Bloch transition, give immediate experimental targets.
Read the full abstract

Computational Complexity

Resilience in labeled real-time automata

In this paper, we characterize resilience for a labeled real-time automaton (LRTA). An LRTA is resilient if whenever a faulty event occurs, after sufficiently many events occur, the LRTA returns to normalcy and the occurrence of the faulty event is not leaked. The notion of resilience reflects the ability of an LRTA recovering from a faulty behavior, and hence can model an intelligent agent. We formulate one definition of resilience for an LRTA and give verification algorithms for the definition based on two basic tools — concurrent composition and observer.
Read the full abstract

Computation and Language

A Hyperbolicity Atlas of Large Language Model Hidden States

LLM hidden states are ordinary vectors, but the distances among those vectors may still show hierarchical structure. To our knowledge, this paper is the first systematic study of whether prompt-token hidden states in contemporary LLMs exhibit Gromov Hyperbolicity (GH), a distance-based measure of tree-likeness. Using 818,904 sample-layer measurements from ten open-weight models across MATH500, HumanEval, WinoGrande, and TruthfulQA, we build a GH map over four axes: parameter scale, layer depth, model family, and input domain. The clearest pattern is depth, not scale: middle layers usually form a high-relative-hyperbolicity plateau, while final layers often become substantially more tree-like. Scale effects are weak and non-monotonic, matched 7/8B model families differ strongly, and domains interact with model specialization. These findings make GH useful as a practical diagnostic: it shows where hierarchical distance structure appears, how specialization changes it, and which model-layer-domain comparisons deserve closer analysis.
Read the full abstract

All authors Zhichao Yang, Yuanze Hu, Gen Li, Qingchen Yu, Shiying Duan, Xinyu Wang, Ye Qiu, Zeming Liu, Guangxu Chen, Zhaoxin Fan

Machine Learning

TrojanWorld: Backdooring World-Model Agents via Imagination Steering

World models increasingly serve as the predictive core of model-based reinforcement learning agents, enabling them to simulate future dynamics and reason over imagined trajectories before acting. Their substantial training demands make pretrained world models attractive for distribution and reuse, exposing downstream systems to model supply chain threats. Backdoor attacks offer a targeted and stealthy means of exploiting such supply chains, yet their threat to interactive world-model agents remains largely unexplored. To fill this gap, we present TrojanWorld, a backdoor framework for world-model agents that induces attacker-specified behavior by steering internal imagination. A physical object placed in the scene acts as the trigger, enabling deployment-time activation through the agent's native observation pipeline without digitally manipulating the observation stream. To achieve effective, stealthy, and persistent control, TrojanWorld combines Decision-Reflective Induction to steer trigger-conditioned imagination toward attacker-specified actions using decision feedback, Clean Behavior Anchoring to preserve trigger-free predictive and behavioral fidelity, and Causal Propagation to sustain the induced preference along subsequent trajectories after the trigger disappears. Together, these mechanisms establish an end-to-end attack chain from physical perception through corrupted imagination to malicious action selection. Experiments with the TD-MPC2, DreamerV3, and R2-Dreamer systems across the DeepMind Control, MetaWorld, MyoSuite, and RoboDesk benchmarks show that under trigger activation, TrojanWorld achieves a target-action deviation as low as 0.026 while retaining at least 98.8% of the corresponding clean performance. Even after trigger removal, the compromised agent can remain trapped in the induced behavioral trajectory, continuing to execute attacker-specified actions.
Read the full abstract

All authors Wenkai Huang, Siyuan Liang, Gaolei Li, Yiming Li, Tianhao Peng, Jianhua Li, Dacheng Tao

Artificial Intelligence

Beyond One-Shot Expansion: Contrastive Evidence Exploration for Multi-Hop Retrieval

Retrieval-augmented generation (RAG) critically depends on retrieving the evidence necessary for effective reasoning. However, this remains particularly challenging in multi-hop question answering (QA), where supporting passages are often linked through intermediate entities and relations that must be progressively uncovered. Existing retrieval approaches typically rely on a single retrieval intent or one-shot query expansion, limiting their ability to adapt to newly retrieved evidence and potentially introducing noisy or redundant retrieval signals. To address these limitations, we propose a training-free multi-hop retrieval framework that integrates evidence-conditioned exploration, passage-specific contrastive refinement, and coverage-aware final ranking. During offline indexing, the framework constructs passage-specific contrastive facets that characterize each passage relative to its semantically similar neighbors, providing fine-grained signals to distinguish closely related candidates. At inference time, the framework iteratively retrieves evidence, generates probes targeting unresolved information needs, refines candidate relevance using the contrastive facets, and selects a complementary set of passages that collectively cover diverse evidence-seeking intents. Experiments on MuSiQue, HotpotQA, and 2WikiMultihopQA demonstrate consistent improvements in retrieval quality and downstream QA performance over baselines.
Read the full abstract

Robotics

Large Discrete Policy: Advancing Explicit Behavior Modeling with Stochastic Iterative Scoring

Behavior policies are often formulated as continuous generative models, whose iterative denoising processes are expressive but difficult to interpret and prone to producing implausible actions. We propose the Large Discrete Policy (LDiP), a fully discrete behavior modeling framework that selects actions from a large vocabulary of physically plausible candidates. Rather than perturbing actions, LDiP improves expressivity through stochastic iterative scoring: it progressively re-scores and prunes candidates with score-space stochasticity, enabling fine-grained ranking and exploration among plausible actions while preserving an explicit decision process. Across end-to-end planning, closed-loop driving, robotic manipulation, and vision-language-action settings, LDiP consistently outperforms strong discrete and continuous baselines in autonomous driving, and exceeds or matches continuous generative policies in robotic manipulation. These results show that discrete policies, when equipped with effective scoring mechanisms, offer an expressive, plausible, and interpretable alternative for behavior modeling. Project website: https://zhenxinli.net/LargeDiscretePolicy/.
Read the full abstract

All authors Zhenxin Li, Nadine Chang, Xinglong Sun, Jingde Chen, Wenhao Yao, Zi Wang, Maying Shen, Yu-Gang Jiang, Zuxuan Wu, Shiyi Lan, Jose M. Alvarez

Computer Vision

FreqDoor: A Hidden Trojan in the Frequency Domain for Backdoor Attacks on Vision-Language Models

Vision-language models (VLMs) have recently shown excellent progress in open-ended image-to-text generation. However, their multimodal nature makes them persistently vulnerable to backdoor attacks. Existing backdoor triggers for VLMs are either spatial, textual, or bimodal, which may yield localized or recognizable trigger patterns. In this work, we explore a different attack surface and propose \ textsc {FreqDoor}, a training-time backdoor attack that implants triggers in the frequency domain. \ textsc {FreqDoor} mixes amplitude-spectrum components from a trigger-source image selectively while preserving the phase of a clean image to generate a spatially distributed and visually imperceptible trigger without modifying the textual input. We evaluate the attack on BLIP-2, InstructBLIP, and LLaVA for image captioning and visual question answering. On Flickr8k, \ textsc {FreqDoor} achieves attack success rates of $99.6\%$, $99.8\%$, and $98.4\%$ on the three models, respectively, while preserving the semantic quality of the generated captions. On VQAv2, the corresponding attack success rates are $99.6\%$, $92.4\%$, and $79.6\%$.
Read the full abstract

Robotics

MEMOBench: A Process Level Memory Benchmark for Robotic Manipulation

Robotic manipulation often requires acting on information that is no longer visible, yet Vision-Language-Action policies are usually evaluated when the current observation largely determines the next action. Existing robotic memory benchmarks expose this gap, but they still rely mainly on final task success and therefore conflate forgetting with manipulation failure. We present MEMOBench, a benchmark for process level memory evaluation in robotic manipulation. MEMOBench includes 30 history dependent tasks, 1{,}500 expert demonstrations, and 4{,}200 executable checkpoint instances from 84 templates. Each checkpoint pairs coarse to fine language with a simulator predicate and labels one memory operation: Storage, Update, or Compression. These annotations define Memory Storage Rate, Memory Update Rate, and Memory Compression Rate, which measure memory fidelity alongside task success. Across standard and memory augmented VLA policies, the strongest memory module baseline reaches only 31.9% average success rate, and high storage often coexists with weak update and compression. Checkpoint language also supervises semantic, contrastive, and framewise memory alignment objectives, yielding modest gains across different memory operations. MEMOBench provides a diagnostic evaluation suite and training supervision for memory grounded robotic policies. The project page is available at https://github.com/Collab-Gen/MEMOBench.
Read the full abstract

All authors Haiyang Sun, Haoxiao Wang, Junming Chen, Weicheng Fang, Zihao Su, Jingkun Yi, Wenyou Yi, Hao Chen, Zhou Zhao

Machine Learning

AI and TCAD for Inverse Design and Defect Discovery: From Simple Machine Learning to LLM

AI has revolutionized various engineering domains, but its impact on semiconductor device design and defect discovery is still limited, due to limited data and the curse of dimensionality. In this paper, we will discuss our work on using the Technology Computer-Aided-Design (TCAD) to generate precise data needed for machine learning (ML) to enable simulation-augmented ML. We demonstrate that with minimal domain expertise, it is possible to create a machine that performs as well as a device engineer on a specific task. We will show that auto-encoder-based machine learning models and noise engineering applied to TCAD data are effective at learning latent physics, and that the models can be seamlessly applied to experimental data. We will demonstrate how to build a device-engineer-level model step by step through various examples, including using only non-destructive electrical data to inverse-engineer the PiN diode layer thickness variations, the Ga2O3 Schottky diode doping and anode workfunction variations, and the transistor contact resistance in an inverter. Examples also include the generation of a FinFET IV/CV prediction model, the mapping between transistor images and IV curves, and the automatic calibration of TCAD parameters for a Ga2O3 Schottky diode, which can only be handled well by experienced TCAD engineers. Finally, to fully realize the potential of AI, large language models (LLMs) and multimodal LLMs (MLLMs) are believed to be necessary. We will discuss the application of LLMs to TCAD command file creation and our vision for MLLMs in automated device design and defect discovery.
Read the full abstract

Optimization and Control

Equivalence Between Continuous-Time Risk-Sensitive Control and Rényi Divergence Minimization

In this study, we show that a continuous-time risk-sensitive control problem is equivalent to a Rényi divergence minimization problem over trajectory path measures. Reformulating stochastic optimal control as probabilistic inference via Kullback-Leibler (KL) divergence minimization avoids the computational intractability of the Hamilton-Jacobi-Bellman equation. However, standard KL control is inherently risk-neutral, and recent minimax extensions remain restricted to risk-averse settings. Our equivalence result resolves this limitation by offering a unified probabilistic framework for arbitrary risk attitudes in continuous-time nonlinear systems. Based on Girsanov theorem, we explicitly map the risk sensitivity to the Rényi divergence order, deriving a noise-dependent control penalty scaled by risk preference. This formulation seamlessly modulates tail-weighting behaviors, interpolating between zero-forcing for risk-averse policies and mass-covering for risk-seeking policies. These findings bridge stochastic control and information-theoretic inference, providing a foundation for sampling-based control algorithms.
Read the full abstract

Analysis of PDEs

Modified Scattering and Asymptotics for Perturbed One-Dimensional Cubic NLS

We study the long-time dynamics of small solutions to the one-dimensional nonlinear Schrödinger equation \[ i\partial_t v+\partial_x^2v-\beta\abs{v}^2v +\cW(x)\abs{v}^4v+\gamma i\partial_xv=0, \] where $\cW$ is spatially localized. The cubic nonlinearity is long range and produces the logarithmic phase correction, whereas the localized quintic term is short range at leading order. We construct a global forward modified wave operator for small complex asymptotic profiles and prove quantitative final-state estimates. For small data in the weighted energy space, we also establish global existence, sharp $t^{-1/2}$ decay, and forward modified scattering with a unique asymptotic profile. The principal new phenomenon occurs beyond this leading law. The exact Duhamel tail generated by the localized quintic term admits a quantitative inner scaling limit at the distinguished frequency $\zeta=-\gamma/2$ on the scale $\abs{\zeta+\gamma/2}\sim t^{-1/2}$. Its universal shape is explicit and depends on the value of the scattering profile on the distinguished ray and on the zeroth moment of $\cW$. When both quantities are nonzero, the limit is nontrivial, belongs optimally to $C^{2,1}_{\mathrm{loc}}$, and is not $C^3$ at the center. Away from the corresponding self-similar ray $\xi=-\gamma$, we construct rigorously defined higher-order outer expansions through every integer order allowed by the decay of $\cW$, and to each fixed finite order when $\cW$ is rapidly decreasing.
Read the full abstract

Cosmology and Nongalactic Astrophysics

Primordial Black Hole Abundances and Scalar Induced Gravitational Waves from Finite-Width Power Spectra in a Stiff Thermal History

We study the formation of primordial black holes (PBHs) from large primordial perturbations that re-enter the horizon during an epoch with equation of state ${\rm w}\geq1/3$. We consider a log-normal curvature power spectrum of finite width $\Delta$ and determine the collapse amplitude by numerical-relativity simulations of a self-gravitating perfect fluid. Threshold scans are performed for five values of $\rm w$ and five spectral widths, and the resulting numerical thresholds are used in the PBH abundance and scalar-induced gravitational-wave (SIGW) calculations. For comparison, we also evaluate the semi-analytical $q$-function prescription. It reproduces the numerical trend close to radiation domination and for nearly monochromatic profiles, but it is not a reliable threshold estimator for generic finite-width profiles in stiffer backgrounds. We show how collapse thresholds increase with both $\Delta$ and $\rm w$, changing the curvature amplitude required for PBHs to constitute all of the dark matter and, consequently, the normalization of the accompanying SIGW signal.
Read the full abstract

All authors Abolhassan Mohammadi, Yogesh, Cristian Joana, Hongwei Tan, Ying-li Zhang

Optics

Modeling of jitter correction methods for asynchronous optical sampling

Timing jitter management is crucial in asynchronous optical sampling (ASOPS) and dual-comb spectroscopy to maintain spectral bandwidth and fidelity. However, a quantitative framework predicting jitter correction efficiency from laser phase noise has been lacking. Here, we present formulation and simulation models that evaluate jitter-suppression dynamics based on phase noise spectra. Our model quantifies the contrast between conventional triggering and software-based jitter correction. It also enables retrieval of a jitter-free spectrum by compensating for jitter-induced power degradation. This work provides a universal design framework for optimizing diverse ASOPS systems by predicting residual jitter prior to system construction.
Read the full abstract