Networking
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Thursday 3 September
Blockchain technology faces scalability challenges because transactions must be validated and recorded across the network. Payment channel networks (PCNs) improve efficiency by moving transactions off-chain and recording only critical interactions on the mainnet. However, PCNs require pre-deposited channel balances (supply) to match transaction demands (demand), and insufficient balances cause supply shortages. Existing approaches, including transaction path optimization and channel balance allocation, address this problem but incur high costs due to dynamic adaptation to fluctuating demands. We reveal a correlation between balance deficits in PCNs and network topology by uncovering the structural organization of liquidity allocation. This enables supply-demand conflicts to be mitigated from a static topological perspective through channel reconfiguration, without additional balance replenishment. We further introduce payment topological entropy (PTE), an information-theoretic metric that quantifies each node's deviation from the global average connection pattern and captures structural properties of balance supply. Based on PTE, we design MaxPTE, a topology optimization algorithm that reorganizes balance allocation across channels through structural reconfiguration, aligning static balance distribution with dynamic transaction demand. Extensive experiments show that MaxPTE reduces balance deficits by 27.25%, increases average maximum flow by 12.23%, and decreases transaction failure probability by 25.96%, outperforming existing benchmarks. The method also remains robust across diverse balance-demand distributions, improving supply-demand balance without prior demand prediction.
Data Analysis and Statistics
Reduced latent leakage does not reliably predict lower likelihood bias in collider inference
Reusable collider representations can be evaluated through downstream discrimination and probes of retained information, but neither quantity directly tests the behaviour of score templates in a profiled likelihood. We test a specific prediction in a controlled two-channel routing protocol: if reduced physics-label readability in a nuisance branch indicates a more inference-robust representation, it should accompany a smaller profiled signal-strength bias under fixed unmodelled shifts. In a public Compact Muon Solenoid $H\rightarrow ZZ\rightarrow4\ell$ workflow, a downstream split of fixed EveNet embeddings preserves signal/background area under the receiver operating characteristic curve ($0.9894\pm0.0004$) while reducing nuisance-branch physics readability from $0.961\pm0.013$ to $0.593\pm0.030$. Probe-sensitivity and effective-rank controls exclude a failed readout and branch collapse. In a separate top quark jet-tagging workflow, the leakage reduction recurs with preserved task performance. Across two development event shards, however, its Spearman association with maximum absolute profiled bias is $0.036$, and three of six material leakage-improving transitions do not reduce that bias. A one-shot preregistered confirmation on an independently accessed shard produces material leakage reductions in all three paired seeds, while the maximum absolute bias increases in two. Thus, within the tested protocol, latent readability is a useful routing diagnostic but not a likelihood-robustness certificate. The result supports a practical validation rule: claims about inference robustness require a prespecified likelihood-facing stress test and held-out confirmation.
Read the full abstract
Performance
RASER: Resilient Agent Scheduling and Execution Runtime for HPC Clusters
The emergence of modern agents powered by large language models has created a demand for executing long-horizon, autonomous workflows in various domains that require significant computational resources. While High Performance Computing clusters provide the ideal infrastructure for these computation-intensive workloads, traditional HPC job schedulers such as Slurm are not designed for dynamic, agentic workflows characterized by unpredictable task durations, external API calls, and fault tolerance requirements of modern agents. This work presents RASER, a user-space framework that enables seamless execution of agentic workflows on production HPC clusters by extending Slurm's internal primitives. RASER introduces agentic job arrays with work stealing via shared filesystem queues, user-space checkpointing through application-level state serialization combined with Slurm requeue, and Apptainer container-based isolation without requiring any image modifications. Evaluations demonstrate that RASER reduces makespan by nearly 39% compared to static partitioning while achieving near-full CPU utilization. RASER provides resilience against preemption and failures while maintaining minimal checkpoint/restore overhead. It requires no kernel privileges or external database infrastructure, making it an accessible solution for deploying agentic workflows on existing HPC infrastructure.
Read the full abstract
Computation and Language
KhatianDoc: A Human-Verified Benchmark Diagnosing Multimodal LLM Failure on Bengali Legal Land Records
Land ownership in Bangladesh is recorded in Ana-Ganda-Kora-Kranti-Til, a base-16 positional fraction system with dedicated Unicode glyphs, no mainstream font, and no coverage in any OCR pipeline or tokenizer. The handwritten records that carry these fractions, RS Khatians, are the authoritative title record for millions of parcels and a frequent subject of civil litigation, yet no benchmark has asked whether a machine can read one. We introduce KhatianDoc, a four-task benchmark built from 107 real RS Khatian records from the Vumi (land) Office of Munshiganj, Bangladesh: symbol recognition, base-16-to-decimal conversion, structured field extraction, and legal document question answering over 1,634 QA pairs. Ground truth was transcribed by hand, verified by a land-law practitioner to full agreement, and anonymized through positional tokens that keep the referential distinctions multi-hop questions depend on. We evaluate six multimodal LLMs (8B to 72B+, open and closed) under a fixed zero-shot protocol. Five QA categories, 39.3% of our stratified set, return zero correct answers from every model; on the arithmetic task, every model that emits a number does worse than a constant-mean baseline, with exact- and near-match scores coinciding: decorrelation, not approximation. Auditing our own metrics surfaced two artifacts in opposite directions: we correct a refusal-scoring bug and report the fixed scores beside the originals, and flag an inflated metadata metric as an upper bound. KhatianDoc documents not a performance gap but the absence of a capability, with verified ground truth for future systems. Code and data, with a redacted image release, are publicly available.
Read the full abstract
Graphics
ReRoom: Blending Virtual and Physical Contexts for In Situ Room Planning in Mixed Reality
Planning a real domestic space is an in situ authoring process: users evaluate candidate layouts at true scale, refine their intent, and carry accepted decisions into later iterations. Existing approaches either separate layout editing from the physical room or provide limited support for evaluating and refining whole-room proposals in situ. We present ReRoom, a mixed-reality system for in situ room-layout authoring. ReRoom presents a shared layout state through a virtual room proxy spatially registered to the target room, allowing interaction and layout generation to remain grounded in the physical context. Users refine the current proposal through direct manipulation or language and preserve accepted placements, allowing each generated update to continue the same evolving design. To balance layout quality with generation efficiency, ReRoom uses a skill-guided layout agent whose room-layout design skill operationalizes three principles that we formulate by synthesizing established interior-design guidance for real-room layout generation. The skill grounds these principles in a normalized representation of the scanned room and reusable geometric checks. Evaluations show that ReRoom produces high-quality layouts for non-rectangular rooms, while its in situ workflow improves the room-planning experience over an otherwise equivalent off-site VR workflow. Code will be released upon acceptance of the paper.
Read the full abstract
Computation and Language
How Far Can Synthetic Data Take Thai OCR?
We investigate what makes synthetic OCR supervision transfer to real Thai documents and use the resulting insights to build Wayu-Paxa-OCR-Zero, a Thai OCR model adapted without OCR labels from real Thai document pages. Synthetic data provide exact labels at scale, but "realism" conflates source domain, page context, typography, spatial structure, and glyph variation. We disentangle these factors with a controlled document-reconstruction pipeline and evaluate each variant under page- and crop-level training on printed and handwritten Thai documents. Non-text context has little consistent effect, whereas typeface diversity, two-dimensional structure, and real handwriting glyphs improve transfer; moreover, source-domain matching depends on training granularity, with in-domain reconstruction approaching real printed supervision under page-level training (1.82% versus 1.31% median character error rate) but underperforming out-of-domain reconstruction under crop-level training (15.59% versus 5.52%). Guided by these findings, we adapt the 0.9B-parameter PaddleOCR-VL-1.6 into Wayu-Paxa-OCR-Zero using 45,723 synthetic pages: relative to its base checkpoint, it reduces median character error rate from 6.64% to 1.24% on printed pages and from 74.87% to 20.55% on handwriting and outperforms Typhoon OCR v1 7B on all five evaluation sets, showing that synthetic-only training can be competitive.
Read the full abstract
Hardware Architecture
LevelSyn: Physical-Aware Logic Synthesis via Level-Asynchronous Graph Neural Networks
As integrated circuit technology scales into the nanometer regime, the traditional disconnect between logic synthesis and physical design has led to significant PPA (Power, Performance, and Area) degradation and prolonged design closure cycles. Traditional logic synthesis relies on non-physical Wire Load Models (WLMs), while recent spectral-based placement predictors often neglect the inherent hierarchical logic depth and signal flow of netlists, which leads to low-fidelity spatial estimations. To bridge this gap, we propose LevelSyn, a novel physical-aware logic synthesis framework that integrates hierarchical representation learning with a wirelength-driven optimization engine. At its core, LevelSyn leverages a level-asynchronous Graph Neural Network (GNN) to predict high-fidelity gate coordinates by capturing the structural and directional semantics of And-Inverter Graphs (AIGs). To handle industrial-scale designs, a level-aligned subgraph partitioning strategy is introduced to eliminate memory bottlenecks while preserving local logical dependencies. These spatial insights are seamlessly integrated into a newly developed physical-informed synthesis engine within the Berkeley ABC framework. Experimental results on the EPFL benchmark suite demonstrate that LevelSyn significantly outperforms state-of-the-art (SOTA) methods, achieving an average power reduction of 6.89% and a timing delay improvement of 27.48%. Furthermore, post-place-and-route validation shows a 99.59% reduction in design rule check (DRC) violations, highlighting its effectiveness in accelerating design convergence.
Read the full abstract
Differential Geometry
New Homogeneous Einstein Metrics from Detection
We use a detection method for invariant Einstein equations to construct and distinguish asymmetric homogeneous Einstein metrics on several families of compact homogeneous spaces. The method retains a geometrically meaningful symmetry-breaking parameter, eliminates the remaining Einstein equations, and then reconstructs positive metrics from the resulting detector. We prove a sharp threshold in a two-factor symplectic family, construct two asymmetric Einstein metrics on each admissible space $\SU(n)^3/\Delta\SO(n)$, obtain a sharp threshold for $\SU(2n)^3/\Delta\Sp(n)$, and construct asymmetric metrics on three exceptional three-factor families. The four-factor case exhibits a further phenomenon. On \[ M_9=\SU(18)^4/\Delta\Sp(9) \] the two symmetry types $3+1$ and $2+2$ are both realized: up to homothety and internal block permutations, the corresponding fixed families contain exactly three asymmetric Einstein metrics, two with three equal horizontal scales and one distinct scale (type $3+1$), and one with two equal pairs of horizontal scales (type $2+2$). They are pairwise non-isometric, Riemannian irreducible, and not naturally reductive. The $2+2$ metric is globally unique in the full positive $2+2$ fixed family up to block interchange.
Read the full abstract
Software Engineering
Code Transformation Rule Synthesis using LLMs: Potential and Limits
Due to their black-box nature, LLMs suffer from limited explain- ability and a lack of determinism. Their usage cost can also rise, particularly with repetitive tasks on large codebases. To mitigate this, we conduct a novel empirical study targeting three domain- specific languages for transformation rules, namely Comby, GritQL, and Ast-Grep. We evaluate three LLMs (GPT-5.4, GPT-oss-120B, and Llama3.1-8B) on six diverse datasets covering four software- evolution tasks: API misuse correction, program repair, API migra- tion, and language version migration. Our results provide evidence that transformation rule synthesis moves beyond proof-of-concept with strong frontier models. GPT-5.4 achieves consistently high rule applicability rates and produces transformations closest to the ground truth across most benchmarks. Smaller and open-weight GPT-oss-120B and Llama3.1-8B models remain effective for simpler, localized changes but struggle with complex migration scenarios. We also observe non-negligible generalizability through the usage of meta-variables and through a high reuse score in the first quartile of many datasets. Finally, when compared to the anti-unification algorithm, LLMs outperform it in correctness, but underperform in rule applicability. Overall, our results show great potential for LLMs to generate sound, correct, generalizable, and reusable rules.
Read the full abstract
Robotics
Scaling Bimanual Household Manipulation from 1,500 hours of Demonstrations to On-Policy Corrections
Learning generalist policies for robust bimanual manipulation is bottlenecked by the scarcity of high quality large scale human demonstration data. In this work, we release 1,500 hours of diverse bimanual manipulation demonstrations covering everyday household tasks, and use this comprehensive corpus to train XR-2, a powerful vision-language-action (VLA) model. Enabled by a purpose built high throughput data pipeline and a carefully designed multi stage training paradigm, XR-2 attains strong manipulation performance in our systematic experiments while retaining favorable training efficiency and high data utilization. We further study two critical scaling axes: varying the amount of expert demonstration data, and post training on DAgger correction data from real time human interventions. In both settings, task success rate improves steadily over the data ranges we probe, exhibiting a clear consistent scaling trend at our current data scale. These results validate both the learning capacity of XR-2 and the promising scaling properties of the released dataset, which we open source to support reproducible research on bimanual robot manipulation learning.
Read the full abstract
Networking
From Prior-Guided Heuristics to Deployable Agents: Accelerating Demonstration-Driven Reinforcement Learning for Deadline-Constrained Network Control
Timely delivery of delay-sensitive information over dynamic, heterogeneous networks is essential for NextG interactive applications, yet providing strict End-to-End (E2E) peak latency guarantees remains an open challenge. Two obstacles limit the adoption of learning-based network control in this setting: traditional volume-based routing metrics, while highly effective for general traffic management, are not designed to capture traffic urgency; and Deep Reinforcement Learning (DRL) controllers trained from scratch suffer from sample inefficiency, long training times, and early-stage exploration volatility. This paper introduces a deployment-focused network control framework that addresses both obstacles. First, we present Effective Congestion (EC), a deadline-aware metric family that quantifies interface congestion by packet urgency and proactively filters non-viable traffic, coupled with a Uniform Path Grouping (UPG) distribution heuristic promoting robust load-balancing; the resulting policies are embedded into Multi-Agent Deep Reinforcement Learning Effective Congestion ($p^*$) (MADRL EC ($p^*$)), a hybrid architecture combining a distributed scheduler with a centralized RL-based router. Second, we introduce a unified training objective that generalizes existing policy-learning paradigms—behavioral cloning, offline Reinforcement Learning (RL), online RL, and offline-to-online schemes—as special cases, combining a live-reward term, a pre-collected-reward term, and a policy-imitation term. From this objective, we derive the Model-Guided Annealed Reinforcement Learning (MGA-RL) protocol, instantiated on a Deep Deterministic Policy Gradient (DDPG) backbone: a deployment-oriented, demonstration-driven training approach that generalizes conventional Offline-to-Online (O2O) schemes, in which trajectories from a lightweight [...]
Read the full abstract
Probability
Correlated initialization of deep residual networks
We study the large-depth behavior of residual networks whose weights are correlated across layers at initialization. Our results confirm and extend a conjecture of Marion et al. [2025], according to which correlated initializations should interpolate continuously between the Brownian stochastic differential equation arising from independent initialization and the ordinary differential equation arising from perfectly correlated initialization. When the initialization is obtained from the application of a feature function to a stationary Gaussian sequence with regularly varying correlation, we prove that there exists a unique critical scaling such that the infinite-depth limit is the solution of a Young differential equation driven by a Hermite process. Hermite processes reduce to the fractional Brownian motion if the feature function generating the initialization has Hermite rank one, which is the case for the identity function, for example. We show that the critical scaling and asymptotic limit are uniquely determined by the decay of correlations together with the Hermite rank of the feature function. Consequently, the correlation structure and Hermite rank of the initialization represent meaningful hyperparameters in the asymptotic regime. By contrast, under finite-variance iid initialization, the asymptotic driver is universally Brownian up to normalization regardless of the choice of distribution. Our proofs rely on a collection of novel results establishing a robust stability theory for Young differential equations in Banach spaces.
Read the full abstract
Artificial Intelligence
KC-Bench: A Dynamic Interactive Benchmark for Evaluating Knowledge Conflicts in LLM Agents
As LLMs increasingly act through tools, they must reconcile user instructions, parametric knowledge, and dynamic environmental observations before taking actions. We introduce KC-Bench, a controlled multi-turn benchmark for measuring this capability across world-knowledge conflicts, input inconsistencies, and multi-source temporal conflicts. Its 238 tasks are manually screened from more than 1,000 generated candidates and combine a user simulator, stateful tools, deterministic environment assertions, an open-source natural-language evaluator, and human trajectory verification. Evaluation of nine models, including DeepSeek-V4-Flash, GLM-5.2, and MiniMax-M3, shows substantial cross-domain variation: no model handles factual correction, identity consistency checking, and temporal conflict resolution reliably across all settings. In the simulated environments, missed conflicts can propagate to tool calls or synthetic protected-data flows. KC-Bench isolates this model-level behavior rather than ranking complete agent frameworks, and provides a reproducible diagnostic for developing conflict-aware reasoning and execution safeguards.
Read the full abstract
Soft Condensed Matter
Effects of Charge Regulation on the Conformations and Dynamics of a Polyelectrolyte during Nanoparticle Adsorption
Customizing the surface characteristics and stimuli-responsive behavior of polyelectrolyte (PE) during interaction with nanoparticles (NP) have potential applications ranging from biological interactions to advanced diagnostic technologies. Using numerical simulations, we investigate how charge regulation (CR) shapes the conformation, dynamics and adsorption of PE on a oppositely charged NP and compared the results with constant charge (CC) conditions. At low solvent ion concentrations, CR barely affects the adsorption fraction, while significantly enhancing the dynamical response of the PE. However, at high ion concentrations, while complexation is suppressed in the CC case, the CR environment helps to overcome the energy barriers related to Debye screening and thermal fluctuations, thereby promoting the adsorption. By maintaining the strong PE-NP electrostatic interaction, PE exhibits a more directed ballistic-like dynamics towards NP during adsorption, in comparison to the diffusive and sub-diffusive behavior observed in the CC case. Our results emphasize that in high ion concentrations, with a strong Debye screening effect, a small fraction of charge-fluctuating sites suffices for adsorption to happen. A tunable CR mechanism is thus an efficient and robust strategy to control the biomolecular interaction and stability in complex environments.
Read the full abstract
Artificial Intelligence
The Attention Triangle in Audio-Video Models
Audio-video diffusion models rely on cross-modal attention to coordinate text, sound, and visual content, yet this same mechanism can introduce subtle and systematic semantic leakage. We study these models by probing and analyzing the “attention triangle,” comprising the three cross-attention edges connecting the text, audio, and video streams, and examine how semantic information is routed across modalities during generation. Our analysis reveals that routing along the audio-video edge is bidirectional: audio can influence video generation, while video can influence audio generation. This edge is shaped by biases encoded in the model's parameters and emerges as a major contributor to leakage: when prompts are in tension with learned priors, cross-modal interactions may override the intended conditioning and reroute semantics toward visually canonical but incorrect outcomes. These effects suggest that semantic artifacts arise not merely from attention spreading beyond its intended target, but from structured, bias-driven interactions along specific pathways. Building on this perspective, we extract attention-derived signals that expose how semantics are distributed and grounded across modalities, and use them as a diagnostic tool to both analyze and deliberately incur leakage under controlled conditions. This enables us to probe the internal dynamics of cross-modal routing and isolate the role of individual interactions. We further leverage these signals to guide inference-time interventions that encourage more consistent cross-modal alignment. Extensive experiments support our analysis and demonstrate improved semantic grounding while preserving generation quality.
Read the full abstract
Computer Vision
Text2Thermal: Physics-Aware Thermal Image Synthesis from Textual Priors
Thermal infrared imaging offers reliable perception in darkness and adverse weather, but thermal datasets remain scarce, motivating extensive work on translating abundant RGB images into thermal. Such translation is fundamentally ill-posed as thermal appearance is governed by surface emissivity and object temperature, neither of which is observable in the visible spectrum, so a single RGB image is consistent with many valid thermal outputs. We argue that language offers a natural means of resolving this ambiguity, and propose **Text2Thermal**, a framework for physics-aware thermal image synthesis from textual priors. Rather than inferring the unobservable radiometric factors from RGB, we supply them explicitly through thermally grounded captions encoding material, weather, time-of-day, and heat-emission state, and adapt a pre-trained Stable Diffusion backbone to the thermal domain. Because the radiometric content is determined entirely by the prompt, Text2Thermal synthesizes thermal imagery without requiring a registered RGB image at inference. Where spatial guidance is desired, an optional control signal imparts scene geometry without disturbing the prompt-specified radiometry. On M3FD and FLIR, Text2Thermal achieves state-of-the-art FID among thermal image synthesis methods, and we additionally report results on the FMB dataset, while offering text-level control that translation-based approaches cannot provide.
Read the full abstract
Sound
Local chord corruption is not recognizer replay: structure-matched calibration
Synthetic chord substitutions offer controlled tests of music generation, but their effects can differ from those of a complete recognized chord sequence. We propose structure-matched calibration, which constructs synthetic chord sequences that match the changed positions and harmonic-relation composition of recognizer replay. Paired generation compares both target-response magnitude and output-chord agreement with replay. On 29 of 30 MUSDB18-HQ songs, central four-second tritone corruption produces a larger target response than complete replay in MIDI-SAG. On 24 held-out MoisesDB songs, structure matching reduces target-response distance to replay by 81% for MIDI-SAG and 77% for MusicGen-Chord. Distance decreases on every song in both models with CNN–CRF. Joint matching also reduces output-chord mismatch with replay by 8–17 percentage points relative to temporal or relational matching alone.
Read the full abstract
Quantum Physics
Optimizing Atom Transport, Gate-Count and Depth with Parity Twine
We present an efficient implementation of the Parity Architecture for neutral-atom quantum processors. We adapt Parity Twine Networks (PTNs) to different atom layouts, native entangling gates, and atom-shuttling capabilities. This provides a general framework for hardware-aware optimization of gate count, circuit depth, and atom transport for quantum circuits encoding arbitrary interaction graphs in a common basis. Specifically, we develop PTN constructions based on different native entangling-gate realizations, namely CZ, CZSWAP, and iSWAP, providing flexibility to accommodate different hardware capabilities on both static and mobile neutral-atom platforms. Using the quantum Fourier transform (QFT) as a representative example, we demonstrate substantial reductions in two-qubit gate count, atom transport, and circuit depth. These resource savings translate into an estimated circuit fidelity three orders of magnitude higher than competing compilation strategies for a 30-qubit QFT. We further extend the construction to the recently introduced optimistic QFT and discuss the broader applicability of PTNs to other quantum algorithms on neutral-atom platforms.
Read the full abstract
Machine Learning
WeatherNext 3: Increasing resolution and performance of global weather models with raw observations
State-of-the-art AI weather models have shown impressive medium-range forecast skill and computational efficiency, but suffer two key shortcomings: their forecasts have lower spatial and temporal resolution than the best physics-based models and they are exclusively initialized with and trained on analysis data. As a result, they cannot directly make use of observations, and any biases in the analysis are inherited by the forecast. WeatherNext 3 addresses these shortcomings and establishes a new state-of-the-art for probabilistic medium-range forecasting skill. First, WeatherNext 3 generates new forecasts every hour (rather than every 6 hours like traditional global models) by ingesting low-latency geostationary satellite data. Second, WeatherNext 3's temporal and spatial resolution are on par with physics-based global models, with hourly time steps and 0.1 degree resolution for single-level variables, including solar radiation and cloud cover. Third, WeatherNext 3 moves beyond traditional analysis variables by learning to predict satellite-derived precipitation estimates, as well as tropical cyclone and station observations. Modelling sparse station data allows WeatherNext 3 to make 2m temperature and dewpoint predictions at any location and time, conditioned on local geographical features, with substantially lower error than competing global models, even when evaluated against unseen stations. Together, WeatherNext 3's capabilities move operational AI-based weather forecasting beyond emulating the traditionally distinct stages of data assimilation, forecasting and post-processing, which helps to further push the frontier of performance and granularity for global weather prediction.
Read the full abstract
Data Analysis and Statistics
The Greedy Bump Bias: Local Profiling Geometry and the Look-Elsewhere Effect
When fitting a localized signal whose position or shape is not known in advance, one typically allows these parameters to vary together with the signal amplitude and chooses the values that maximize the likelihood. This freedom has two related statistical consequences. If a genuine signal is present, its fitted amplitude will be affected by a positive bias; otherwise, the same freedom increases the chance of finding an unusually signal-like background fluctuation, giving rise to the look-elsewhere effect. We show that these two effects can be understood as consequences of the same local geometry of the family of signal templates. We study this connection in a Gaussian matched-filter model, where a smooth D-dimensional family of normalized templates describes the unknown signal location or shape. In the normalized matched-filter problem, the curvature of a genuine signal peak and the fluctuations that determine the curvature of a high background peak are governed by the same template metric. This allows us to derive an explicit asymptotic relation. We then follow the problem away from the strong-signal and high-threshold limits. Separating the signal-associated maximum from the best competing maximum gives an exact decomposition of the global bias into a local profiling contribution and a contribution from remote-peak competition. In a one-dimensional Gaussian example, the second factorial cumulant accounts for most of this correction, while the third brings the prediction into close agreement with simulation. A two-point Kac–Rice calculation reproduces the second cumulant and reveals a quartic short-distance suppression of nearby maxima. The resulting picture separates the roles of local dimension, model-dependent curvature, and global extremal competition within a common framework.
Read the full abstract
Artificial Intelligence
HalluPeer: A Taxonomy-driven Benchmark for Detecting Hallucinations in Scientific Peer Reviews
The growing scale of academic peer review has motivated the use of Large Language Models (LLMs) as review assistants, yet LLMs can generate fluent but unsupported claims that undermine review reliability. Existing hallucination benchmarks are not designed for peer review, where verification requires grounding claims in long, technical papers. We introduce HalluPeer, a benchmark for detecting hallucinations in scientific peer reviews, providing aligned triples of paper content, human-written reviews, and hallucination-injected reviews, annotated for detection, classification, and localization. Our pipeline induces a peer-review-specific hallucination taxonomy, identifies review contexts, and injects hallucinations with automated filtering. Experiments on 12K papers and 38K reviews show that existing detectors struggle to separate hallucinations from legitimate critique, while evaluation on authentic reviews demonstrates that HalluPeer-defined hallucination patterns occur in real peer reviews, highlighting the critical need for source-aware verification. Our project page can be found in https://github.com/Lin-TzuLing/HalluPeer.git
Read the full abstract
Nuclear Experiment
First Measurement of Near-Threshold J/{\psi} Photoproduction on the Neutron
We report the first measurement of the total and t-differential cross sections for near-threshold J/$\psi$ photoproduction on the neutron, obtained with the CLAS12 detector at the Thomas Jefferson National Accelerator Facility. The measurements, performed using a liquid-deuterium target, also provide the incoherent total and t-differential cross sections on the bound proton, enabling the first direct comparison of near-threshold J/$\psi$ photoproduction on bound protons and neutrons. Interpreted within Vector Meson Dominance, holographic QCD, and GPD-based frameworks, the data allow for the investigation of the gluonic structure of bound nucleons. A comparison with free proton results reveals hints of modifications to the gluon structure of nucleons in the nuclear medium, within the assumptions of the extraction and on the underlying production mechanism. The comparison of bound neutron and bound proton cross sections provides new constraints on the near-threshold J/$\psi$ production mechanism, which is crucial in order to establish J/$\psi$ photoproduction as a probe of the nucleon's gluonic structure.
Read the full abstract
Materials Science
High-Throughput Computational Discovery of Inverted Resistive Switching in Two-Dimensional Materials
Atomristors, non-volatile resistive switching devices based on two-dimensional (2D) monolayers, are promising building blocks for energy-efficient memory and neuromorphic computing. However, their design remains restricted to a few materials such as MoS2 and h-BN, limiting functional diversity and design flexibility. Here, a high-throughput computational framework combining density functional theory, machine-learning molecular dynamics, and quantum transport simulations screens about 2,900 exfoliable monolayers for vacancy-mediated resistive switching, identifying 17 thermally stable candidates in two mechanistically distinct classes. In Class 1 monolayers, such as GaS, Au adsorption at the native vacancy introduces conducting states, switching the insulating monolayer from a high- to a low-resistance state (HRS-to-LRS). Class 2 monolayers, comprising ionically bonded metal oxyhalides and nitrohalides such as BiOCl, exhibit previously unreported inverted switching. Vacancy-released electrons delocalize and push the Fermi level into the conduction band, placing the device natively in the LRS; Au adsorption re-localizes these carriers and returns the Fermi level to the gap, driving LRS-to-HRS switching. Quantum transport simulations confirm both mechanisms, while migration-barrier calculations identify the electrode-2D separation as a key parameter governing Au migration and the resistance window. These findings expand the atomristor landscape and establish complementary switching as a design paradigm for multifunctional memory and neuromorphic hardware.
Read the full abstract
Computation and Language
Language, Language Models, and What We're Talking About
Language models are commonly discussed as technical artefacts, but they are obviously shaped by the linguistic worlds conveyed by data during their training. Using Italian language models as evidence, I want to bring attention to the nature of the systems which result from training and specialising models on translated and synthetic data, and further curating them, and to the meaning of testing them on equally unnatural data. Are these eventually models of Italian? Are they models of language? Does NLP still care about language? These questions yield another, more concrete question: what language do we actually want language models to produce? I argue that this question cannot be answered if we do not first consider a clearer distinction between language models designed as technical products and language models designed as tools for studying language itself. The answers then might be diverse, the languages we are talking about might be diverse, and the picture might not be as pessimistic as we fear.
Read the full abstract
Materials Science
SALTED: a symmetry-adapted machine-learning program for predicting electron-densities in molecules and materials
SALTED provides an open-source Python package for machine learning the quantum-mechanical electron density, $n(\mathbf{r})$, in molecular and condensed-phase systems based on input atomic coordinates and species. The program adopts a linear atom-centered decomposition of the electron density, which makes it highly transferable across diverse atomistic configurations sharing similar chemical environments. Because of this representation choice, SALTED is naturally interfaced with state-of-the-art electronic-structure programs based on atomic orbitals, namely CP2K, FHI-aims, and PySCF, from which reference electron-density data can be generated and used to train a model. The learning algorithm is based on a symmetry-adapted extension of Gaussian process regression, making SALTED especially efficient in small-data regimes. Thanks to the implementation of vector-field kernel functions, SALTED can also learn the first-order response of the electron density to applied electric fields, $\partial n(\mathbf{r})/\partial \mathbf{E}$. The application of SALTED within computational workflows has already shown its utility in a wide variety of contexts, including the calculation of polarization vectors and polarizability tensors, the accurate evaluation of Coulomb forces in QM/MM molecular-dynamics simulations, and electronic-structure studies of large-scale 2D materials.
Read the full abstract
Earth and Planetary Astrophysics
Trans-Neptunian Object dynamics even better explained by a stellar flyby after 4.5 Gyr of evolution
The Trans-Neptunian objects (TNOs) formed together with the planets from a flat disc of gas and dust but now orbit the Sun mostly on inclined, eccentric orbits. One explanation for the TNOs' orbits is a close flyby of another star. One with a mass of $M_p =$0.8 Msun at $q_p =$ 110 au and $i_p =$70{\deg} fits the observations. However, such close encounters were more frequent when the Sun was young and still part of its birth cluster. Assuming that the flyby happened then, we use numerical $N$-body simulations to model how the TNOs' orbits changed due to interactions with the planets over the 4.56 Gyr since the Sun formed. We find that the cold Kuiper belt region lost about 80% and the hot Kuiper belt population about 40% of its initial population. About 7% –8% of the TNOs were injected into the planet region, but almost all (99%) were ejected afterwards. By contrast, the orbital parameters of distant Sedna-like objects ($q >$60 au) remained basically unchanged. Surprisingly, in sum, these changes improve the fit to the observed TNO population even more, strengthening the argument for a close flyby to the Solar System. The remaining differences concern the location of the cold population and the inclination distribution of the Sedna population. We discuss possible reasons and steps to resolve these issues.
Read the full abstract
Data Structures and Algorithms
An $m+n^{3/2}$ algorithm for counting spanning trees by $\ell_1$-regularized resistance
We study the basic problem of approximating the number of spanning trees of a graph. For a graph with $n$ vertices, $m$ edges, We propose an algorithm that approximates the number of spanning trees in $\widetilde O(m+n^{3/2}\eps^{-1})$ time. Our algorithm improves upon the previously best known $\widetilde O(m+n^{15/8}\eps^{-7/4})$ time algorithm by Chu, Gao, Peng, Sachdeva, Sawlani, and Wang [FOCS 2018] and the $\widetilde O(m^{3/2}\eps^{-1})$ time algorithm by Liu, Peng, and Yang [FOCS 2026]. Notably, our algorithm is based on the novel concept of $\ell_1$-regularized resistance. We propose simple and efficient algorithm for computing $\ell_1$-regularized resistance and we show that they can be used to approximate the number of spanning trees by combining with the determinant sparsifier framework of Durfee, Peebles, Peng, and Rao [FOCS 2017]. Our algorithm matches the best known size of the determinant sparsifiers.
Read the full abstract
Dynamical Systems
Diffeomorphism Group Actions on Para-Kähler Structures: Lifts and Induced Dynamics on 3-Webs
Let $M$ be a smooth manifold equipped with a para-Kähler structure. We define an action of the diffeomorphism group $\operatorname{Diff}(M)$ on the space of para-Kähler structures and prove that this action preserves the associated affine structures. Although this result is well known, we provide a proof based on this action, showing in particular that an isometry commutes with the Levi–Civita connection. We then construct several natural liftings of this action to the tangent bundle $TM$, the cotangent bundle $T^*M$, and the Whitney sum $E=TM\oplus T^*M$. In certain cases, these lifted actions induce natural dynamical systems on a distinguished class of $3$-webs.
Read the full abstract
Computer Vision
Drive-HWM: Hierarchical World Models for Dynamic-Latent Guided Autonomous Driving
World models offer a promising paradigm for autonomous driving by predicting how traffic scenes may evolve and using such predictions to support action generation. However, existing approaches either separate future prediction from action generation or jointly predict them at the same temporal scale, making it difficult to simultaneously achieve long-horizon anticipation and responsive, observation-grounded decision making. We present Drive-HWM, a hierarchical slow–fast world modeling framework that organizes future representation prediction and action generation at complementary temporal scales. The slow world model predicts multi-step future representations to capture extended scene evolution. To explicitly model the abundant motion dynamics in driving environments, we introduce Dynamic-Aware Latents learned through optical-flow prediction. Guided by these future representations, the fast model uses a lightweight multimodal backbone and an autoregressive expert to jointly predict the next frame and the immediate action from the latest observation. Next-frame prediction encourages the fast model to capture imminent scene evolution, while one-step action generation allows decisions to be continuously updated as new observations arrive. Extensive experiments on NAVSIM v1 and v2 demonstrate the strong driving performance of Drive-HWM. Comprehensive ablation studies further validate the effectiveness of the hierarchical slow–fast design, dynamics-aware future representations, and joint next-frame and action prediction.
Read the full abstract
Soft Condensed Matter
Heterogeneous collapse in thermoresponsive copolymer microgels varying molar composition
Understanding the internal architecture of copolymer microgels is crucial for establishing how nanoscale polymer organization controls their stimuli-responsive behavior. Here we focus on thermoresponsive P(N-isopropylacrylamide-co-N-isopropyl-methacrylamide), P(NIPAM-co-NIPMAM), microgels with varying mole fraction of the components, synthesized via radical precipitation polymerization, and we demonstrate that changes in their volume phase transition and equilibrium swelling are governed by composition-dependent internal heterogeneity. Comparison between small-angle neutron scattering (SANS) with isotopic labeling and monomer-resolved simulations show a block-like monomer distribution of the two components. SANS analysis reveals a universal maximum in the polymer mesh correlation length near the transition, evidencing coexistence of collapsed NIPAM-rich and swollen domains. The correlation length increases with increasing NIPMAM content, with a maximum for 75 mol % NIPMAM, implying sparse collapsed regions within the network and thus a large degree of heterogeneity induced by the presence of an increasingly large fraction of intercalated, non-collapsing PNIPMAM. The maximum heterogeneity correlates with the equilibrium swelling ratio, indicating that collapsed microgels retain a structural memory of the transition and present a less-compliant structure in response to temperature variations. Overall, these insights highlight a complex effect of the block-like monomer distribution on the responsive properties of copolymer microgels with different compositions, thus providing a design rule for tailoring responsive colloids for functional soft materials.