Quantum Physics
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Monday 7 September
We study a single-particle Lindblad lattice in which nonreciprocal incoherent hopping is combined with a tunable coherent boundary link. The link changes the Hamiltonian boundary condition while the incoherent jump rates are kept fixed, so it does not directly alter the diagonal population generator. Its effect on the Liouvillian gap comes from population-coherence-population feedback: the Hamiltonian couples populations to damped coherences, and eliminating those coherences produces a frequency-dependent self-energy for the slow population branch. For the periodic reference system, where translation symmetry is imposed on both the Hamiltonian and the dissipator, we derive an exact fixed-momentum reduction, a scalar secular equation for the population-connected branch, and the associated hydrodynamic drift and coherence-corrected diffusion. For open finite lattices, exact diagonalization and sparse near-zero eigensolvers show that the coherent boundary link can enhance or suppress the gap in one dimension and produce geometry-dependent responses in two dimensions. We also show that an observed relaxation scale can differ from $1/\Delta_{\mathcal L}$ when the gap family has weak visibility in the chosen perturbation and observable. The periodic construction gives exact analytical benchmarks, while finite-size simulations in one and two dimensions demonstrate how the same coherence-mediated mechanism controls open-boundary gaps and observable relaxation.
Probability
The $L^q$-bounds for Derivatives of Unitary Developments of Random Continuous Geometric Rough Paths
In this paper we derive an explicit formula for derivatives of unitary developments of random continuous geometric rough paths of all orders and establish proper $L^q$-bounds for them, which allow us to study the analyticity of their characteristic functions in a quantitative manner and then prove that under some mild conditions the distributions of the signatures of random continuous geometric rough paths are determined by their expected signatures, provided the latter have positive radius of convergence. In particular, we give a partial affirmative answer to an open question posed by T. Lyons and H. Ni in “Expected signature of Brownian motion up to the first exit time from a bounded domain” (The Annals of Probability, 43(5), 2729-2762, 2015), concerning whether the expected signature of stopped Brownian motion determines the law of its signature.
Read the full abstract
Accelerator Physics
Higher-order space-charge stability in anisotropic beams: Vlasov-Poisson derivation, refined dispersion relations, and stability charts
The Hofmann stability chart is used to screen working points in space-charge-dominated linacs. We identify two errors in its published higher-order dispersion relations: missing $(1\mp2\hat\eta^2/\alpha)$ factors in the third-order $S^4$ coupling residues, and a sign error in the stated isotropic reduction of the fourth-order relation. Both corrections follow from Hofmann's Vlasov-Poisson equations without fitted parameters. They reproduce coherent tune-shift coefficients in the author's later monograph that the printed forms miss by 24% and 127%. Mode-resolved figures from a published application agree with the corrected relations and reject the printed forms, indicating an inconsistency between the 1998 equations and the calculations underlying those tested figures. We quantify the effect on the non-oscillatory stability chart. Inside the adopted $S^2\le10$ comparison domain, printed and corrected forms disagree on 0.73-2.11% of cells, with no preferred direction. Among excluded cells, disagreement reaches 22%, and the printed relation over-predicts instability at every sampled anisotropy. This concentration may help explain why the errors persisted, although it does not establish their historical cause. For PIP-II, the corrected chart flags four of thirty-two evaluable periods, including one on a third-order odd branch missed by a second-order screen. This count covers non-oscillatory modes only and remains conditional on an unresolved factor-five disagreement between two codes on transverse emittance growth.
Read the full abstract
Computer Vision
AnomalyCraft-700K: Component-Level Controllable and Verifiable Synthetic Anomalies for Fine-Grained Video Anomaly Understanding
Progress in video anomaly understanding (VAU) has long been limited by inherent deficiencies of real-world anomaly videos, which are hard to collect and offer little control over their content. Synthetic anomaly approaches partially alleviate data scarcity, yet their generation remains largely controlled at the category or prompt level. They also lack component-level verification of video-text consistency and provide insufficient hard normal samples near the normal-anomaly boundary. To address this, we present AnomalyCraft-700K, a component-controllable synthetic anomaly dataset for fine-grained VAU, containing over 40K videos and over 700K task-level textual annotations. From fine-grained semantic components and a progressive three-stage pipeline, we craft anomaly events that are richly detailed, semantically controlled, and temporally structured, and additionally construct per-category hard normal samples to prompt the model to discriminate based on anomaly semantics rather than surface visual cues. Moreover, using the components as verification units, AnomalyCraft-700K further performs component-wise correction of video-text discrepancies introduced during generation, providing reliable annotations with verified cross-modal alignment for six tasks that progress from anomaly detection, through anomaly retrieval and captioning, to fine-grained anomaly reasoning. Evaluations of widely used methods under both traditional and MLLM-based protocols demonstrate that AnomalyCraft-700K serves as an effective source of supervision, from anomaly detection to fine-grained anomaly understanding.
Read the full abstract
General Relativity and Quantum Cosmology
A New Window for Testing the PPN in the Strong-Field Regime: Stellar Rotation of S301
The recently discovered S301 star, with an orbital period of 8.7 years and eccentricity $e = 0.982$, reaches a pericenter of just $140\,R_{\mathrm{s}}$ — approximately nine times closer than S2's $\sim 1200R_{\mathrm{s}}$ — offering an unprecedented laboratory for testing strong-field gravity. We demonstrate that S301's stellar spin precession provides a novel probe of the PPN parameter $\gamma$, scaling as $(2\gamma_{\mathrm{PPN}} + 1)/3$ relative to general relativity (GR). For maximum projected velocity shifts of $46.1~\mathrm{km\,s^{-1}}$, next-generation spectrographs could constrain $\gamma_{\mathrm{PPN}}$ to $\sim 10\%$ precision. When combined with S301's orbital precession — a substantial $\sim 1.95^\circ$ pericenter advance per orbit (GR prediction), approximately an order of magnitude larger than S2's $\sim 0.2^\circ$, constraining $(2\beta_{\mathrm{PPN}} + 2\gamma_{\mathrm{PPN}} - 1)/3$ — the spin measurement breaks the $\beta$-$\gamma$ degeneracy. A MCMC analysis of projected S2 and S301 observations yields $\sigma_\beta \approx 2.3 \times 10^{-4}$, with the $\gamma$ precision limited by the spin precession measurement to $\sigma_\gamma \approx 0.1$. While the spin precession provides an independent consistency check, its true significance lies in probing gravity at $\phi/c^2 \sim 10^{-4}$ and $v/c \sim 0.08$ — four orders of magnitude stronger than Solar System tests and an order of magnitude stronger than current S-star constraints. This complementary approach opens a new window for testing the PPN framework in the strong-field regime and demonstrates the power of combining multiple stellar probes at the Galactic Center.
Read the full abstract
Machine Learning
HealthLoopQA: A Context-Aware Question Answering Benchmark for Interpreting Wearable Monitoring Data in Diabetes Care
As medical wearables become integrated into daily chronic disease care, effectively interpreting longitudinal monitoring data is essential for patients and clinicians to understand health trends, detect safety-critical events, and make informed decisions. While large language models (LLMs) show promise for transforming this streaming physiological data into personalized health insights, evaluating their reasoning capability and analytical rigor in diverse monitoring tasks remains a fundamental challenge. Existing medical wearable question answering (QA) benchmarks primarily assess short-horizon classification or statistical summaries, largely ignoring the long-term patterns, therapeutic and behavioural contexts, and potential system failures inherent in real-world deployments. To address this, we introduce HealthLoopQA, a comprehensive diagnostic benchmark for evaluating LLM reasoning over continuous diabetes monitoring data. Grounded in a novel taxonomy of eleven atomic reasoning abilities, HealthLoopQA comprises 127 tasks and over 1,500 QA instances spanning process mining, anomaly detection, and prediction over 30-day horizons. To systematically evaluate safety awareness, we complement real-world datasets with a fault-injected simulation testbed modeling diverse device malfunctions and cyber-physical attacks to generate physiologically plausible hazard scenarios. Evaluating state-of-the-art LLMs across prompting and agentic frameworks reveals severe limitations in complex temporal pattern mining. Furthermore, we identify a broader phenomenon of In-context Laziness under long-context prompting, highlighting critical open challenges in deploying LLMs for rigorous long-horizon medical reasoning.
Read the full abstract
Number Theory
Maximising the number of solutions to linear equations
We study the asymptotically maximal possible number of integer solutions to the linear equation $ax+by+cz = 0$ with a fixed choice of $a, b, c \in \mathbb{Z}$ and variables $x, y, z \in S$ for some finite set $S\subset \mathbb{Z}$, as $|S|\to +\infty$. Define $\gamma_{a, b, c}$ to be the largest constant for which there are arbitrary large finite sets $S\subset \mathbb{Z}$ such that the number of solutions to $ax+by+cz=0$ with $x,y,z\in S$ is $\gamma_{a,b,c}|S|^2-o(|S|^2)$. We prove structural results for general $a, b, c$ and moreover, we show that $5/13\le \gamma_{1,1,-3}\le 1/2-\delta$ for some constant $\delta>0$. In addition we show that the limit as $a \rightarrow \infty$ of $\gamma_{1,1,-a}$ is equal to precisely $1/5$.
Read the full abstract
Computation and Language
Train Overcomplete, Deploy Compact: Scaling Recovery Capacity for Structured LLM Pruning
Large language models achieve strong performance across diverse tasks, but deployment remains costly because of memory, latency, and energy demands. Structured pruning reduces these costs by removing architectural components, yet its recovery stage is often limited by a mismatch between the recovery module's representational capacity and the complexity of the removed knowledge. We call this bottleneck the capacity-knowledge asymmetry and propose OverRep, an Overcomplete Reparameterization framework for structured LLM pruning. Following the principle of "train overcomplete, deploy compact", OverRep temporarily overparameterizes the recovery module during training to absorb complex knowledge distilled from the original model. After recovery, the overcomplete re-parameterization is algebraically merged into a mathematically equivalent compact module, preserving the pruned model's inference-time architecture and computational cost. OverRep further introduces an annealed activation that enables nonlinear training dynamics while converging to a linear regime for exact algebraic merging. Across three backbone families, OverRep improves retained reasoning performance over strong recovery baselines by up to 5.5 and 8.4 points at 25% and 50% pruning, respectively, while keeping memory usage and TFLOPs comparable to existing recovery methods. Our code is available at https://github.com/mmai-laboratory/OverRep.
Read the full abstract
Cryptography and Security
Lightweight Detection of Electromagnetic Signal Injection Attacks on Image Sensors
Electromagnetic signal injection attacks (ESIA) pose a growing threat to image sensors, which are increasingly used in different intelligent systems. By emitting electromagnetic interference, adversaries can manipulate pixel values, potentially misleading downstream artificial intelligence (AI) models and causing unsafe decisions in these systems. We present a lightweight detection method that leverages optically black pixels, which are non-exposed pixels already present in many modern image sensors, to identify the attacks. Our detection approach achieves an area under the receiver operating characteristic curve (ROC-AUC) of up to 99.6% and an Equal Error Rate (EER) as low as 0.027 across diverse attack conditions. Our method requires minimal computational overhead and no hardware modifications, making it a practical and effective defense for securing vision-based systems against ESIA.
Read the full abstract
Cryptography and Security
AgentDrift: A Step-Labeled Benchmark of Injection-Hijacked LLM Agent Trajectories
LLM agents complete tasks by issuing sequences of tool calls, and every observation they read is a channel through which an indirect prompt injection can enter. A successful injection has a characteristic shape when the trajectory is read in order: a benign prefix gives way to actions that serve the attacker rather than the user. Existing benchmarks measure whether such attacks succeed against live agents, and existing guard models judge a trace as a whole; no public corpus labels, step by step, where an injection enters a trajectory and which steps it corrupts. We present AgentDrift, a benchmark of 12,536 synthetic tool-call trajectories over five agent domains in which every one of the 71,024 steps carries one of four labels: benign, injection point, hijacked, or failed injection. The corpus contains 4,000 benign, 5,536 attacked, 1,500 failed-attack, and 1,500 hard-negative trajectories; attacked trajectories follow three compliance patterns whose label strings obey a stated regular grammar. Failed attacks carry an injection the agent resisted, and hard negatives carry legitimate content that resembles an attack, so a detector must separate attempt from success and deviation from novelty. Trajectories were generated by a single open model under category-specific protocols, enforced by a closed-vocabulary structural validator, screened by an LLM judge, and audited by hand on 1,200 trajectories; we show that the LLM judge was itself fooled by the hard negatives. A surface-feature logistic regression recovers only 55.4% of attacks (F1 0.647), including only 8.2% of partial hijacks and 23.1% of delayed executions, so nearly half of the attacks require modeling the behavioral sequence. We measure template concentration, attack-goal-family collapse, and world-identity leakage in the generated data, and release the corpus with its documentation under CC BY 4.0.
Read the full abstract
Information Theory
Equivalence of Fixed-Rank and Rank-One Even-Order Symmetric Tensor Factorization
In the recent work of Barbier, Ko, and the second present author on sublinear-rank symmetric matrix factorization [Math. Stat. Learn. 9 (2026), 1-68], a key result is that, in the Bayes-optimal setting, the large-size limit of the free entropy of the finite-rank spiked Wigner model is the same as in the rank-one case when the signal has centered i.i.d. entries. In this paper, we show that this rank-one equivalence result extends to the case of finite-rank, even-order, symmetric tensor factorization. Moreover, we give a natural reformulation of a hypothesis that was stated in the aforementioned work to be necessary for this result. As in the matrix case, we use information-theoretic identities and replica symmetry to reduce a known multi-dimensional variational formula for the limiting free entropy to its one-dimensional analog. The novelty stems from the fact that said formula involves a replica symmetric potential containing Hadamard (entrywise) powers, rather than squares, of the matrix-valued variational parameter, so the eigenvalue-based approach used in the matrix case must be adjusted.
Read the full abstract
Computation and Language
CantoneseLLM v2: Reasoning in a Low-Resource Language
Cantonese is widely spoken but remains low-resource in written data, with no large corpus of native Cantonese reasoning traces available for model training. We develop and release CantoneseLLM v2, comprising models based on Qwen3 8B and 30B-A3B. The models are trained through CPT on 784 million Cantonese and Hong Kong-related tokens, chat-vector merging, SFT, DPO, and RLVR. Evaluation across the training stages shows that chat-vector merging transfers instruction following but preserves the donor model's reasoning language, while SFT with limited Cantonese reasoning data substantially shortens or removes reasoning traces and reduces benchmark performance. DPO restores the reasoning-block format, particularly for the 8B model, but recovers only part of the lost performance. The RLVR training with Cantonese language and Traditional Chinese scripts as multiplicative constraints introduced Cantonese language alignment and restored the lost performance. The 30B-A3B model reaches 73.16 on HKCanto-Eval, within 1.20 points of its merged checkpoint, while retaining the Cantonese reasoning behaviour absent from that checkpoint. We release the model checkpoints, the training environments, and a thirteen-year Traditional Chinese Common Crawl dataset. The models can be accessed at https://huggingface.co/collections/hon9kon9ize/cantonesellm-v20
Read the full abstract
Signal Processing
Networked Embodied Communication: From Collective Distinguishability to Communication Reliability
Embodied agents need to convey information to surrounding infrastructure, but their active communication interfaces may be unavailable, constrained, or intentionally inactive. Their ability to manipulate physical states offers a complementary path: messages can be encoded in deliberately selected configurations and recovered through infrastructure sensing. This principle underlies embodied communication. Yet physical differences do not guarantee distinguishable messages: a single sensing viewpoint may leave ambiguities that repeated sensing cannot resolve. This paper develops networked embodied communication, where distributed access points (APs) jointly observe message-bearing scatterer positions under fixed illumination. Under a correlated Gaussian sensing model, we characterize the additional distinguishability supplied by receive APs, establish exact redundancy conditions, and reveal how distinctions absent from individual observations can emerge through cross-AP statistical relationships. We then establish the exact asymptotic optimal maximum-error behavior of a finite alphabet under repeated independent sensing. The largest group of indistinguishable messages determines the error floor; once all messages are distinguishable, the minimum pairwise Chernoff information determines the error exponent. For a given alphabet, receiver cooperation can therefore eliminate an error floor that repetition at any individual AP cannot overcome. Building on these results, we derive finite-budget reliability conditions and jointly design the receive AP set and message-bearing positions. Numerical results show that the proposed search closely approaches exact benchmarks on reduced instances with substantially fewer candidate evaluations than exhaustive enumeration, while receiver cooperation reduces the sensing intervals needed to guarantee reliable decoding.
Read the full abstract
Information Retrieval
Tracing Query Expansion Effects through Sparse Autoencoder Features
Query expansion (QE) is a critical technique in information retrieval that enriches underspecified queries with additional textual context. However, its effect is often unreliable in modern dense retrieval, especially for strong off-the-shelf retrievers without retraining. Existing studies mainly examine expansion quality, semantic drift, or retrieval outcomes, but rarely explain how QE changes dense retrievers internally. In this work, we trace QE effects through sparse autoencoder (SAE) features. Using paired original and expanded queries, we decompose layer-wise retriever representations into sparse latent activations, identify QE-related latents from expansion-induced activation shifts, and interpret them with natural-language descriptions and retrieval cases. Our analysis shows that effective QE induces layer-concentrated changes in sparse latents aligned with retrieval intent and entity attributes, rather than only perturbing final query embeddings. SAE-based activation steering further validates these latents improve retrieval more consistently than random interventions or vanilla QE across four benchmarks, suggesting that SAEs can explain QE effects and offer a lightweight option for precise retrieval behavior modulation without query rewriting or retriever fine-tuning.
Read the full abstract
Computer Vision
Re-calibrated Contrastive Loss for Transformation-Aware Prompt Conditioning in Vision-Language Models
Ensuring effective transfer learning for vision-language models without compromising their generalization performance is crucial. However, many existing methods overlook data characteristics and simply reuse the training strategies adopted during pre-training. Specifically, they treat same-class samples as distinct instances and transform images independently of their paired text prompts, which makes model learning more difficult. We address these limitations through transformation-aware prompt conditioning and a re-calibrated contrastive loss. Fixed text descriptors identify the transformations applied to paired images, providing transformation-level consistency without altering class semantics. This design aligns the image and text branches at the transformation level, enabling richer representations while preserving the models' ability to generalize. In addition, our loss function mitigates positive-gradient dilution in soft-target cross-entropy when each anchor has multiple valid positives. During transfer, our approach treats same-class samples as positives rather than distinct instances, enabling the model to learn domain-specific features more effectively. Experiments across distribution shift, transfer learning, and few-shot settings demonstrate consistent improvements over existing approaches. Source code for our method is available at https://github.com/SoongE/ReCalCon.
Read the full abstract
Machine Learning
MOLE: Detecting Insider Threats in AI Agents
Model misalignment, prompt injection, or operator misuse could lead AI agents operating frontier-lab accounts to exfiltrate model weights, poison training data, or weaken release gates. Existing benchmarks do not test whether defenders can detect this activity among routine work under a limited review budget. We introduce MOLE, an open benchmark of 150 AI-operated accounts sharing 9 stateful services over 30 workdays, with 12 threats and 8 corpora from four models totaling roughly 20 billion tokens. Of 39 agent models, 72% complete most assigned harmful objectives and agent refusal does not predict completion. MOLE enables comparison of 40 monitors across corpus generators, observability levels, and threats; even the best evaluated monitor in our single-day audit-event comparison misses nearly half of completed harm. MOLE also enables monitor development: benchmark-guided search improves a mid-tier monitor by 49-64%, while selective use of a stronger monitor improves budget-AUC by 10% over applying it to every account-day at comparable modeled cost.
Read the full abstract
Numerical Analysis
Null-Space-Free 6D Spectral Embedding with Local Rayleigh Quotient Recovery for 3D Quasiperiodic Maxwell's Eigenproblems
We develop a numerical framework for three-dimensional quasiperiodic Maxwell eigenvalue problems obtained through a six-dimensional periodic embedding. A projected Bloch–Fourier discretization yields a structured generalized eigenvalue problem with a large gradient-type kernel. Explicit orthonormal bases for the longitudinal and transverse subspaces remove this kernel exactly and reduce the original generalized eigenvalue problem to a null-space-free standard eigenvalue problem containing only the positive spectrum. An explicit inverse representation of the reduced operator avoids nested inner–outer linear solves and leads to an inverse Lanczos method whose main inner computation is a Hermitian positive definite conjugate-gradient solve with condition number bounded by that of the mass matrix; a residual estimate quantifies the effect of inner solves on the inverse Ritz pairs. To recover the computed modes in physical space, the six-dimensional Fourier eigenvectors are reconstructed on a three-dimensional Yee grid by a separated multi-center Taylor expansion, which avoids the dense Fourier-to-grid phase matrix and remains practical when direct dense reconstruction becomes prohibitively expensive. Local weighted Rayleigh quotients provide an independent physical-space validation, and their mass-weighted expectation is proved to equal the cropped Yee Rayleigh quotient under a partition-of-unity condition. Numerical experiments confirm the accuracy and computational effectiveness of the proposed framework and its physical-space recovery of three-dimensional quasiperiodic Maxwell modes.
Read the full abstract
Information Retrieval
FunnelAudit: Responsibility Auditing in Multi-Route Recommender Systems
Multi-route recommender systems combine retrieval, allocation, fusion, and ranking, making individual inclusions and exclusions difficult to audit. Route overlap can hide effects from one-at-a-time ablations, while freezing downstream stages produces counterfactuals inconsistent with serving behavior. We introduce FunnelAudit, an executable framework for incident-level responsibility auditing. An accountability contract specifies the disputed Top-K event, controls and owners, permitted reference actions, and replay semantics. FunnelAudit evaluates every permitted control configuration and applies graded actual responsibility to find the smallest outcome-preserving contingency that makes each control pivotal. Its certificate records the contingency and paired serving executions needed to verify the judgment. We instantiate the framework in two-stage, nine-route funnels using fixed union, weighted quota allocation, or weighted reciprocal-rank fusion, followed by SASRec ranking. Across 258,809 user-target incidents from three real interaction datasets, 4.24-16.24% admit a responsible control. Among responsible incident-control pairs, 92.55-99.64% require a nonempty contingency, so single-control ablation recovers only 0.36-7.45%. Policies differing in factual outcomes on only 0.31-2.39% of incidents yield 21.44-54.05% Jaccard distance between responsible-route sets on matched exclusions. Independent replay reproduces all 9,121,792 checked target-world outcomes; exhaustive search and a generic mixed-integer linear program agree with every sampled judgment. These findings demonstrate the importance of explicit serving semantics and checkable witnesses for recommender accountability.
Read the full abstract
Computation and Language
TurEngMix: A Text Corpus and Benchmark for Turkish-English Code-Mixed Language Identification and Named Entity Recognition
Natural language processing systems underperform on code-mixed text, particularly for low-resource language pairs. Turkish-English poses a further challenge: it lets English stems combine with Turkish suffixes to form single mixed-language tokens. We introduce TurEngMix, a corpus of 5.5K noisy, naturally occurring social media posts (486,974 tokens) rich in Turkish-English code-mixing. From this corpus, we construct a new Turkish-English benchmark for code-mixed language identification (LID) and named entity recognition (NER), comprising 15K expert-annotated tokens. Evaluating both decoder LLM and fine-tuned encoder baselines, we find that monolingual Turkish and English tokens are labeled reliably, but all models have high error rates on mixed-language tokens for both LID and NER. For morphologically integrated tokens, NER error rates were 5.2x and 6.3x higher for GPT-4o and Qwen, respectively. This highlights how morphological integration remains a challenge. We release the corpus, annotations, and code to support future computational and sociolinguistic research on Turkish-English code-mixing.
Read the full abstract
Computer Vision
DPSF-Net: A Dual-Prior Spatial-Frequency Network for Real-World Remote Sensing Image Dehazing
Real-world remote sensing image dehazing (RSID) remains challenging because atmospheric scattering, spatially non-uniform haze and colour distortion jointly degrade structural and spectral information. Most deep learning methods rely on RGB inputs and spatial-domain feature extraction, which limits their ability to separate global background haze from local surface details. Here, we propose DPSF-Net, a dual-prior spatial-frequency network built on MCAF-Net for real-world RSID. The network uses hazy RGB images and dark channel prior (DCP) maps as joint inputs, allowing physical degradation cues to guide end-to-end feature learning. A spatial-frequency residual interaction block introduces a FourierUnit branch into multi-directional spatial interaction to model large-scale haze components. A prior-guided feature attention module adaptively fuses prior and attention features to reduce colour shift and structural distortion. A selective kernel complementary fusion module screens multi-scale skip features through bidirectional residual complementary gating and selective kernel fusion. Extensive experiments demonstrate that DPSF-Net achieves state-of-the-art performance on the real-world RRSHID remote sensing image dehazing benchmark and remains competitive across multiple synthetic datasets. Moreover, the proposed method strikes a favourable balance among restoration quality, parameter count and computational complexity, supporting the effectiveness of dual-prior spatial-frequency modelling.
Read the full abstract
Artificial Intelligence
SSP-DMGTimeNet: Physics-Constrained Learning for Spatiotemporal Trajectory Prediction of Vehicle Platoons
Existing car-following prediction methods mainly optimize trajectory accuracy, while rarely considering whether predicted disturbances propagate realistically along a vehicle platoon. This limitation may lead to accurate but string-unstable predictions. We propose SSP-DMGTimeNet, a physics-constrained learning framework for spatiotemporal trajectory prediction of vehicle platoons. The model combines multi-scale temporal representations with cross-vehicle interaction features to capture complex and time-varying platoon dynamics. A propagation-delay-aware causal attention mechanism explicitly models upstream-to-downstream disturbance propagation by learning response delays between adjacent vehicles and accumulating them along the platoon. In addition, time- and frequency-domain string-stability losses relieve disturbance amplification across both adjacent vehicles and arbitrary sub-platoons during training. Experiments on HighD show that SSP-DMGTimeNet achieves an unstable-window rate of 0.65% for five-vehicle platoons and a maximum head-to-tail amplification of 0.898 on the ground-truth excitation subset, while maintaining competitive trajectory prediction performance. In zero-shot evaluation on NGSIM US-101 and I-80, the model achieves velocity MAEs of 1.316~m/s and 1.252~m/s, with unstable-window rates of 3.90% and 4.10%, respectively. These results demonstrate that incorporating platoon-level physical constraints can effectively balance trajectory prediction accuracy and disturbance propagation stability.
Read the full abstract
Artificial Intelligence
iBrain: A Unified Foundation Model Reading the Brain from Surface to Spikes
Invasive neural recordings provide high-fidelity measurements of brain activity, with signals such as intracranial EEG (iEEG) and intracortical spiking activity capturing neural dynamics at different spatial and temporal scales. Yet existing neural foundation models have largely been developed independently for different invasive recording paradigms, leaving joint pretraining across heterogeneous invasive signals underexplored. In this work, we introduce iBrain, a unified foundation model that jointly learns from iEEG and spiking activity. iBrain employs signal-specific encoders to accommodate their distinct signal characteristics and a shared spatiotemporal Transformer backbone to model dependencies across recording channels and time. We pretrain iBrain on over 7,000 hours of heterogeneous neural recordings using masked signal reconstruction and channel-view alignment, promoting contextual modeling of neural dynamics and robustness across different channels. iBrain consistently outperforms single-signal pretraining baselines and achieves state-of-the-art performance on multiple benchmarks. Further experiments demonstrate that iBrain exhibits transferability and data efficiency across diverse recording settings. These results highlight the potential of joint pretraining on heterogeneous invasive neural recordings to support scalable neural modeling and transferable representations across recording settings and downstream tasks.
Read the full abstract
Computer Vision
MSSP: Multi-Scale Spatially-Constrained Partition for Unsupervised Semantic Segmentation of 3D Point Clouds
3D point cloud semantic segmentation is essential for real-world spatial understanding, yet the prohibitive cost of human annotations motivates unsupervised approaches that require no labels. Existing superpoint-based methods typically rely on spectral analysis at a fixed granularity, failing to capture the hierarchical semantic structures inherent in complex indoor scenes. To bridge this gap, we present a Multi-Scale Spatially-Constrained Partition (MSSP) framework that combines multi-scale spectral analysis with spatially-constrained clustering. Multi-scale spectral analysis constructs enriched superpoint descriptors across multiple clustering granularities; however, the resulting high-dimensional feature space calls for a structural prior to translate into cleaner segmentation. Spatially-constrained clustering supplies this prior by restricting superpoint merging to physically adjacent regions, imposing the spatial coherence needed for multi-scale features to be effective. Extensive experiments on S3DIS and ScanNet show that MSSP achieves the best mIoU among unsupervised methods on the main benchmarks, with particularly significant gains on S3DIS. Notably, our ablation reveals a regularize-then-enrich interaction: multi-scale features alone do not improve final segmentation, yet become highly effective when coupled with spatial regularization, underscoring that spatial coherence is aprerequisite for multi-scale representations in superpoint clustering.
Read the full abstract
Robotics
Mind the Phase: Effective Rank and Representation Health in Legged Locomotion
Reinforcement learning has become the leading paradigm in legged locomotion, enabling complex behaviors from backflips to parkour via massively parallel simulation. Under PPO's non-stationarity, shallow networks remain the de facto architecture, supported by carefully staged environments; nevertheless, the representations learned by these policies remain poorly understood, specifically the relation to motion produced on hardware. In this work, we empirically study locomotion policies through the effective rank of the network Jacobian and show that conditioning rank on gait phase exposes representation structures invisible from a global rank. In particular, we find that architectural choices, namely observation and layer normalization with residual connections, allocate more than one additional dimension of effective rank to swing-dominant states than to stance-dominant states, which is mostly absent in vanilla MLPs. Building on this, we propose practical recommendations to convert these into diagnostics for healthier and smoother policies. In practice, this results in lower joint jitter, up to $2$x, not only in simulation, but also on a physical Spot. More broadly, our findings show that how a policy distributes its capacity across the gait cycle is closely tied to its health and sim-to-real transfer.
Read the full abstract
Quantum Physics
Trace-class spectra of irreducible Gaussian quantum Markov semigroups
We determine the spectrum of the predual generator of an irreducible Gaussian quantum Markov semigroup on trace-class operators, the natural space for density operators. The drift matrix governs the linear part of the Gaussian evolution and organizes the spectral classification. When the linear flow generated by this matrix is not periodic, the spectrum is the closed left half-plane. If all drift eigenvalues have negative real part, every interior point is an eigenvalue. There are no trace-class eigenvectors if some drift eigenvalue has positive real part or all have zero real part. For periodic flow, the spectrum consists of unions of horizontal half-lines or filled parabolic regions. Examples show that irreducibility cannot in general be omitted. We apply the results to oscillators, free particles, and boson chains.
Read the full abstract
High Energy Physics (Experiment)
Neutron detector response modeling in NOvA
Neutrons can present a significant challenge for neutrino experiments in which energy reconstruction is critical. With the ability to escape detection completely and with a weak correlation between their kinetic energy and any eventual energy deposition, it is difficult to fully account for neutrons produced in neutrino interactions. This in turn leads to significant model dependence when evaluating neutron-related systematic uncertainties. The NOvA experiment is a long-baseline neutrino oscillation experiment with a high-statistics sample of antineutrino data collected by its near detector. We report an excess relative to data of simulated neutron candidates with low energy depositions when using standard Geant4 physics lists. The simulation excess is traced to an overabundance of secondary photons produced from interactions of neutrons with kinetic energy greater than \SI{20}{\mega\eV}. Improved agreement with data is obtained by applying the data-driven neutron-on-carbon \menate model for neutrons between \SI{20}{\mega\eV} and ${\sim}$\SI{100}{\mega\eV}. With \menate, the residual oversimulation is more uniform across the calorimetric neutron energy spectrum, suggesting possible overproduction of primary neutrons by the GENIE neutrino interaction generator. These results motivate the adoption of \menate-supplemented Geant4 simulation as the nominal simulation in the production of future \nova simulation.
Read the full abstract
Computer Vision
Joint-Conditioned Stereo Surface Reasoning for Interaction Field Estimation
Predicting hand–object interaction fields requires locating the nearest object-surface point for each hand joint, often from small and partially occluded image regions. We view this task as joint-conditioned surface-endpoint estimation: each joint has its own nearest endpoint, while endpoints from the same hand can draw on shared local surface evidence. This structure motivates Joint-Conditioned Stereo Surface Reasoning (JSSR). A temporal-stereo network jointly predicts 3D joints, a direct interaction field, and per-view endpoint evidence. Calibrated candidate search evaluates endpoint hypotheses using joint-specific image compatibility and cross-view correspondence. A hand-shared candidate support lets joints draw on common surface evidence, and a learned residual gate controls the geometric correction when observations are ambiguous. Our system built on this method ranked third on the SHOW3D Interaction Field Challenge leaderboard.
Read the full abstract
General Relativity and Quantum Cosmology
The Dynamical Instability of Rotating Boson Stars
We investigate the dynamical instability of rotating boson stars described by the Gross–Pitaevskii–Poisson equations with contact self-interactions. Through three-dimensional simulations, we confirm that the rotating boson star undergoes a quasiperiodic conversion between ring-like and twin-star-like density configurations in the early nonlinear stage. We develop a systematic linear stability analysis to identify the modes driving this instability and show that repulsive self-interactions could significantly increase the lifetime of the rotating boson stars. We further construct a three-mode Hamiltonian to describe the early nonlinear stage, which explains the quasiperiodic conversion. This analytical framework agrees reasonably well with the simulation results and provides a clear picture to understand the dynamics of rotating boson stars.
Read the full abstract
Game Theory
Deception in Reach-Avoid Game with Unknown Heterogeneous Attackers Speed Information
This letter investigates a reach-avoid game involving two Attackers and one Defender, where the Attackers aim to maximize the number reaching the target region while the Defender seeks to minimize it. In contrast to conventional complete information formulations, we consider an information asymmetry scenario where the Attackers' heterogeneous maximum speeds are privately known but publicly disclosed to lie within continuous ranges. Existing studies on uncertain speeds, however, have primarily focused on homogeneous settings, whereas heterogeneity extends the uncertainty from a common capability level to the relative capability configuration of the Attackers. To address the resulting capture-order ambiguity over infinitely many possible speed combinations, we establish a critical speed pair framework that characterizes when different capability configurations induce different optimal capture orders, and enables the analysis of the Defender's guessing behavior and the design of information-limiting strategies for the Attackers. We demonstrate that under certain initial conditions, the Attackers can mislead the Defender into making suboptimal decisions through a slow-speed deception strategy, achieving superior payoffs compared to the complete information game. Numerical visualizations reveal the widespread occurrence of such dilemma conditions.
Read the full abstract
Optics
Self-powered InAs nanowire detector arrays for extended-SWIR spectrometry at room temperature
Spectral sensing in the extended shortwave infrared (e-SWIR) is important for molecular analysis, infrared imaging, and machine vision, motivating the development of compact spectrometers for broader applications. However, conventional commercial off-the-shelf spectrometers in this wavelength region are expensive and bulky due to their reliance on external dispersive optics/filters and/or cryogenic accessories. Other emerging computational spectrometers are based on Si and InGaAs photodetectors that remain focused on the visible and near-infrared, with few detector platforms operating in the e-SWIR regime that simultaneously provide broadband sensitivity, low-noise room-temperature operation, and diverse spectral signatures for accurate identification and reconstruction. Here, we report a room-temperature e-SWIR computational spectrometer based on InAs/InP core-shell nanowire photodetector arrays with geometry-encoded spectral responses. The detectors exhibit self-powered broadband photoresponse across the 1–3 $\mu$m range, with responsivity up to 0.215 A W$^{-1}$, detectivity up to $1.6 \times 10^{9}$ cm Hz$^{1/2}$ W$^{-1}$, and microsecond response times. The excellent detector performance is leveraged to demonstrate filter-free spectral reconstruction using a compact multipixel photodetector array device. This enables high-accuracy molecular absorption spectrum reconstruction and hyperspectral imaging. Our results indicate that InAs nanowire arrays are a promising platform for compact computational spectrometry and imaging in the e-SWIR at room temperature.