Peer Revenue Publish or profit

Papers

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

All of arXiv

Sunday 6 September

Number Theory

A Fricke transformation for cubic-residue eta products of level 13

We prove an explicit Fricke transformation for the two-dimensional space generated by normalized reciprocal products on the cubic-residue cosets modulo 13. The transformation matrix is a scalar multiple of a matrix of differences of cubic Gaussian periods. Its projective action is defined over the cyclic cubic field, whereas the normalized matrix is defined over the real cyclotomic field. The proof uses generalized eta functions, a modular unit of degree two on $X_1(13)$, and a five-coefficient identity. A separate divisor argument lifts the projective transformation to the asserted linear transformation. We also record the Fricke images at an arbitrary prime $p\equiv1\pmod3$ and prove that their normalized sine constants belong to the associated cyclic cubic subfield.
Read the full abstract

Logic

Preserving the Ultrapower Axiom by Forcing over Nonoverlapping Extender Models

We identify a class of simple yet nontrivial forcing notions that preserve the Ultrapower Axiom (UA) over canonical inner models, reaching the level of a strong cardinal. The forcings we consider are discrete product forcings whose index sets are chosen and spaced according to the structure of the underlying nonoverlapping extender sequence. As an application, we show that UA is consistent with the existence of a strong cardinal $\kappa$ such that $V\neq \text{HOD}_X$ for every $X\in V_\kappa$, answering a question of Goldberg.
Read the full abstract

Quantum Physics

Spectral Twisting in a Common Bosonic Reservoir: Fragility of Two-Qubit Dark-State Protection

The interaction of two qubits with a common bosonic reservoir is encoded by a matrix-valued spectral density $\mathbf{J}(\omega)$, whose diagonal entries describe the local spectra, while the off-diagonal entries encode cross-correlations. Even when \(\mathbf{J}(\omega)\) has rank one, a frequency-independent dark channel need not exist because the family \(\{\mathbf{J}(\omega)\}_{\omega}\) may have a trivial common kernel. We term the frequency-dependent rotation of the bright and dark directions spectral twisting and quantify it through the Fubini–Study speed $\tau(\omega)$ of the bright spectral projector. We analyze how twisting modifies two-qubit dynamics and quantify the loss of dark-state protection through the leakage \(P_{\mathrm{leak}}(t)\). Comparisons with untwisted asymmetric reservoirs and rotating-wave dynamics, together with detuned and finite-temperature calculations, distinguish spectral twisting from coupling asymmetry, counter-rotating processes, and thermal absorption. For resonant qubits tuned to the crossing of the two local spectra $\omega_\times$, the singlet is locally dark at the transition frequency but couples to off-resonant components whose bright directions are rotated. In the weak-twisting regime, the fixed-time leakage scales as $P_{\mathrm{leak}}(t)\propto[\omega_\times\tau(\omega_\times)]^2$. We test this prediction for mismatched Drude-Lorentz spectra using nonperturbative hierarchical equations of motion generalized to cross-correlated bath forces. These results provide a geometric framework for dark-state engineering in structured reservoirs.
Read the full abstract

Networking

Enabling Emergency Communication via Semantic Radar-Centric ISAC

Combat aircraft and other modern military platforms field two co-located RF assets: a radar for sensing and a dedicated radio for communication. In contested battlefield environments, however, the dedicated communication path is highly likely to be targeted and disrupted by jamming, blockage, or physical damage, while the radar remains fully operational. A natural fallback is to repurpose the surviving radar aperture as an emergency communication channel. The difficulty is that radar is purposely engineered for sensing rather than communication, so loading any information onto its waveform perturbs the beampattern and inevitably erodes sensing performance: the more bits we push, the more sensing we lose. We identify the core problem as enabling emergency communication via a radar-centric integrated sensing and communication (ISAC) link while preserving the radar's primary sensing function. Therefore, we propose a semantic radar-centric ISAC framework that pairs a learned vector-quantized semantic encoder, which compresses task-relevant content into a few tens to a few hundred bits, with a wideband phased-MIMO radar restricted to a conservative spatial-spectral configuration subset that respects a bounded sensing-degradation budget. Experiments on SAR target classification and text classification show that our framework supports reliable emergency communication for downstream tasks while keeping the radar beampattern almost unchanged.
Read the full abstract

All authors Mohaimin Al Barat, Chaoyu Zhang, Hexuan Yu, Kai Zeng, Y. Thomas Hou, Wenjing Lou

Quantum Physics

Hilbert space geometry change by a nonadiabatic electromagnetic field

A change in the geometry of the Hilbert space of a quantum system interacting with a nonadiabatic electromagnetic field is found. To identify its physical origin, the evolution of orthogonality to non-orthogonality of the states of the quantum system has been investigated based on the matrix elements of the quantum states in three main physical cases: a quantum system in zero external field (closed quantum system), a quantum system in slowly changing (adiabatic) field, and a quantum system in rapidly changing (nonadiabatic) field. The quantum states are orthogonal in the first two cases and change to non-orthogonal in a nonadiabatic field, following continuously the parameters of the field. A time-dependent Fubini-Study metric of the Hilbert space can be determined based on the matrix elements of the quantum states in the nonadiabatic case. Such a dynamical metric of the Hilbert space in quantum mechanics resembles the dynamical metric of spacetime in general relativity.
Read the full abstract

Quantum Physics

Critical Non-Hermitian Skin Effect and Scale-Free Localization Morphing in Bilocally Coupled Hatano-Nelson Chains

We investigate the critical non-Hermitian skin effect (CNHSE) and the coupling-driven morphing of scale-free localization (SFL) in two Hatano-Nelson chains coupled locally at two bulk sites separated by an inclusive distance d. At weak interchain coupling, competition between the opposite skin accumulations of the two chains generates a geometrically selected critical branch with a complex loop-like spectrum and scale-free skin states localized near the physical boundaries. Remarkably, this SFL branch persists as the coupling becomes strong, although its underlying mechanism and spatial profile change qualitatively. Strong local hybridization generates high-energy impurity states and effectively divides the low-energy Hilbert space into geometry-selected inner and outer segments, with the former supporting SFL states near the interchain-coupled sites and the latter retaining conventional skin states. The number and localization scale of the SFL states are controlled by the separation between the coupling sites. Consequently, increasing the interchain coupling continuously relocates the SFL states from the physical boundaries to the coupling-defined internal interfaces. Our results demonstrate that bilocal coupling can induce a CNHSE and further drive its scale-free states into a segmentation-induced localization regime.
Read the full abstract

Information Retrieval

Who Anchors AI Overviews in Health? Baidu, Google, and the Geography of Authority

Artificial intelligence is being rapidly incorporated into traditional search systems, yet scant work audits the information disparities across platforms, geography, and languages. We address this gap by comparing Google and Baidu's AI Overview systems for health queries, and measure informational anchors that emerge. Auditing 1,920 health queries across 12 countries and 4 languages, we find that Google and Baidu exhibit vertical integration, routing users toward their own company platforms in AI Overviews rather than a diverse set of primary sources. Smaller, lower-localization countries receive fewer domestically sourced references for health queries. Issuing the same query in a country's official language rather than English raises the share of locally sourced citations approximately 3.5- to 13.5-fold. Comparing queries across health topics of varying severity and controversy, including Traditional Chinese Medicine as an example, we also show that health disclaimers are multidimensional and vary across language and culture. We discuss how generative search influences access to health information, and the urgent need for culturally-aware oversight of these systems that influence critical health decisions.
Read the full abstract

Logic

Approximation by simple structures

In her graduate thesis, Gwyneth Harrison-Shermoen developed an abstract framework for approximating structures of a given theory, which generalized the notion of smoothly approximable structures. At the time it was unclear how broadly the framework applied due to a lack of examples. In this paper we will show that with slight modification to the framework, several well-understood $\mathrm{NSOP}_1$ theories are in fact approximable by simple structures. We discuss certain structural properties of the limit structure that can be obtained from the approximation and introduce a notion of limit dimension as a candidate for recovering Kim-independence in the limit structure. We also provide some examples of structures that are not approximable by structures of finite Morley rank.
Read the full abstract

Hardware Architecture

Hardware Trojan Threats to Multi-Chiplet Photonic Neural Network Accelerators

Multi-chiplet photonic neural network accelerators (MCPNAs) combine photonic computation, photonic communica-tion, and heterogeneous chiplet integration to enable scalable and energy-efficient AI acceleration. However, their distributed archi-tecture and reliance on third-party chiplets introduce significant hardware security risks. This paper examines Hardware Trojan (HT) threats to MCPNAs across three dimensions: confidentiality, integrity, and availability.
Read the full abstract

Number Theory

Generalized Zeckendorf expansions of the 2nd order

We generalize Zeckendorf's theorem to second-order linear recurrences of the form $G_{k+2} = gG_{k+1} + hG_{k}$ with arbitrary, coprime initial values $(G_1, G_2)$. We analyze the asymptotic behavior of the count $\#R_G(X)$ of positive integers less than or equal to $X$ that have an expansion under the standard rule of expansions $\mathcal{E}$ associated with the recurrence. Additionally, we provide sufficient conditions under which $G$ has the unique expansion property under the rule of expansions. Finally, we completely characterize the $g$-golden ratio recurrence family when $G_1 = 1$ or $G_2 = 1$. Our approach directly investigates the algebraic structure of the expansions, simplifying and completing previous partial results.
Read the full abstract

Geophysics

Hourly U.S.-wide flood simulation beyond the limits of traditional and data-driven models

As increasingly-damaging floods can strike within hours of a storm and in ungauged reaches, hourly network-wide simulation has become critical societal infrastructure. Here we demonstrate a multi-timescale physics-embedded learning model which outperforms the United States' operational system and surpasses AI-based systems at flood peaks. Covering more than 800,000 river reaches of the conterminous U.S., the model elevates median hourly Nash-Sutcliffe efficiency at 2,831 gauges to 0.683 from 0.461 for the operational National Water Model v3.0, and narrows flood-peak timing errors from 7-8 hours to 4.5-6 hours. dHBV2.0MTS-MC captures 33% more >=50-year floods than NWM3.0 and 159% more than an operational LSTM baseline. Against recent AI models, overall hourly skill is comparable while rare-flood accuracy is distinctly higher, with relative peak-magnitude error reduced by 34% for >=100-year floods. It combines long-term hydrologic context, short-term shocks, and infiltration excess to resolve extraordinary hourly peaks not visible on a daily plot. Process-model parameters and hourly discharge are produced for every reach, seamlessly covering the continent at 7.2 km2 median resolution. This candidate for the next-generation National Water Model sets a new operational accuracy level for national-scale flood prediction.
Read the full abstract

All authors Wencong Yang, Leo Lonzarich, Yalan Song, Haoyu Ji, Ming Pan, Kathryn Lawson, Chaopeng Shen

Representation Theory

A short note on Arthur parameters with singular orbit closures

This short article resolves a question about Arthur parameters highlighted in previous work of Balodis, Cunningham, Fiori and Zhang, in the case of $\mathrm{GL}_n/F$ where $F/\mathbb{Q}_p$ is a $p$-adic field. Namely, we prove that every Arthur parameter $\psi$ of $\mathrm{GL}_n/F$, can be factored into Arthur parameters $\psi\cong \psi_1\oplus\cdots \oplus\psi_r$ such that the orbit $C_\psi=C_{\psi_1}\times\cdots\times C_{\psi_r}\subseteq V_{\lambda_{\psi_1}}\times \cdots \times V_{\lambda_{\psi_r}}$ in the associated Vogan variety has smooth closure if and only if each $\bar{C}_{\psi_i}$ is smooth for $i=1, \ldots, r$, and that $C_{\psi_i}$ has smooth closure if and only if it is open or closed.
Read the full abstract

Artificial Intelligence

Improving Proficiency and Efficiency of Android GUI Agents via Self-Generating Tool Actions

Android agents using a hybrid action space that combines GUI actions and tool actions (e.g., accessing application data via APIs) remain largely underexplored, mainly due to the excessive effort required to create tools. To address this gap, we introduce DroidTool, a framework for augmenting the agents with self-generated tools, which are realized as Python functions operating on application states (e.g., a database). To create tools with minimal human labor, DroidTool employs an agentic workflow featuring stages: proposal, implementation, test generation and execution, and repair. Notably, when testing the created tools for verification, it constructs relational tests across relevant tools for natural preparation of appropriate test preconditions and improved test coverage, rather than testing each tool separately. The GUI agents augmented with the generated tools achieved approximately 4.47%p higher performance with approximately 20.05% fewer interactions than the GUI-only agents, averaged across representative benchmarks: AndroidWorld, B-MoCA, and MobileSafetyBench.
Read the full abstract

Cryptography and Security

Efficient Hardware Information-Flow Tracking for Pre-Silicon Security Testing

Register-Transfer Level (RTL) simulation is widely used to test hardware before it is fabricated. To allow testing for security related information flow properties, such as confidentiality and integrity, taint logic can be automatically added to the design to track how information flows through it. However, taint logic instrumented by the state-of-the-art, such as CellIFT, makes simulation-based testing prohibitively expensive: On our evaluation of Mega-BOOM (136K cells), it increases the instrumented design to 5.81x the original cell count and causes a 143.72x simulation slowdown. The taint logic could be simplified to improve simulation speed, but it will inevitably trade off its precision. This lightweight, imprecise taint logic will introduce false positives and may eventually result in even more overhead to check these false positives. This paper explores the research question of where precision is actually needed in the design to overcome the overhead of false positives. It presents CEGAR-T, a framework that automatically synthesizes taint logic that minimizes the taint-logic instrumentation overhead while guaranteeing no false positives (relative to the precise CellIFT baseline). We have implemented CEGAR-T and evaluated it on the safe instruction set problem for timing side-channel security across open-source RISC-V cores. Over all evaluated cores, CEGAR-T reduces both instrumentation and simulation overhead, in geometric-mean, from 5.64x to 1.42x and from 34.65x to 1.79x, respectively, without compromising the precision benefit of the CellIFT baseline.
Read the full abstract

All authors Yu-Wei Fan, Yuheng Yang, Christine Guo, SooHyuk Cho, Thomas Bourgeat, Mengjia Yan, Sharad Malik

Signal Processing

Tri-Hybrid Beamforming Design for DMA-Aided Secure ISAC Systems

This paper proposes a tri-hybrid beamforming scheme for secure integrated sensing and communication (ISAC) with a dynamic metasurface antenna (DMA) architecture, where the base station (BS) is capable of communicating with legitimate users and sensing the target. There is also an eavesdropper in the system intending to eavesdrop on the confidential information. The tri-hybrid beamforming design problem is formulated with the objective of maximizing the sensing signal-to-noise ratio (SNR) under the constraints of secrecy spectral efficiency (SSE), transmit power, and physical structure limitations. We first solve the problem and obtain the optimized fully-digital beamforming solution through successive convex approximation (SCA) and semidefinite relaxation (SDR) approaches. A triple alternating optimization scheme is then developed to iteratively optimize the digital, analog, and DMA beamformers, progressively approximating the fully-digital solution. Numerical results demonstrate that the proposed secure tri-hybrid beamforming design for DMA-aided ISAC improves the sensing SNR by approximately 3 dB compared to a tri-hybrid beamforming scheme with a fixed DMA electromagnetic design.
Read the full abstract

All authors Siyi Li, Zhuoming Li, Mohammadali Mohammadi, Jiajun He, Hien Quoc Ngo, Michail Matthaiou

Information Theory

Improved Degree Tables for Secure Distributed Matrix Multiplication

In secure distributed matrix multiplication, a user wishes to compute the product of two matrices with the assistance of servers, in such a way that any $T$ colluding servers learn nothing about either matrix. Degree tables are a combinatorial tool for constructing polynomial codes for this problem and underlie several state-of-the-art schemes, including $\mathsf{GASP}_r$, $\mathsf{GASP}_{r,s}$, and $\mathsf{DOG}_{r,s}$. We introduce a periodic-gap framework for constructing degree tables that contains these three families as special cases and leads to two new constructions, $\mathsf{SHIFT}_{r,s}$ and $\mathsf{COVER}_r$. We determine their exact recovery thresholds and show that, in many cases, they outperform the current state of the art. We also prove a new lower bound on the recovery threshold of ordinary integer degree tables. In the balanced case, in which the partitioning parameters and the security parameter are equal, we sharpen this bound to match $\mathsf{SHIFT}_{r,s}$ up to lower-order terms, showing that it is asymptotically optimal among degree-table constructions.
Read the full abstract

Computer Vision

When Speech Meets Lips: Interpretable Audio-Visual Synchronization for L2 Pronunciation Assessment

Automatic Pronunciation Assessment (APA) systems have achieved strong performance with transformer-based models and self-supervised speech representations. However, most methods rely only on acoustic signals and overlook temporal synchronization between speech and articulatory movements, limiting diagnostic feedback on timing mismatches important for L2 pronunciation training. We propose an interpretable audio-visual synchronization framework that explicitly models speech-lip temporal alignment through feature encoding, cross-attention fusion, lag estimation, stability quantification, and visualization. The framework introduces frame-level lag trajectories and a Lag Stability Index (LSI) to quantify synchronization robustness. We also interviewed 30 participants, including 10 instructors and 20 students with diverse first-language backgrounds, to assess its effectiveness. By transforming implicit alignment into interpretable representations, the framework connects automatic scoring with actionable Computer-Aided Pronunciation Training feedback. Datasets and supplemental materials are available at https://www.robots.ox.ac.uk/~vgg/data/lip_reading/.
Read the full abstract

All authors Bowen Yu, Mingyu Huang, Yishen Liu, Yue Zhao

Atmospheric and Oceanic Physics

A scaling theory for the macroturbulence of weakly supercritical planetary atmospheres

A theory for the general circulation of the atmosphere must be based on a theory of its macroturbulence. A central component is the near-surface eddy heat flux, which is closely tied to the mass transport of the surface branch of the circulation. Existing scaling theories are largely formulated within the two-layer quasi-geostrophic framework, which assumes supercritical states and an inverse cascade of kinetic energy—conditions that are frequently not satisfied in planetary atmospheres, including Earth's. Nevertheless, such theories have shown some empirical success, including that the Rhines scale is associated with the mixing length even in the absence of any inverse energy cascade. Here, we analyze hundreds of idealized, dry general circulation model simulations spanning wide ranges of rotation rates, meridional temperature gradients, vertical stratification, and seasonality. Most simulations reside in the marginally critical regime where strongly nonlinear scaling theories fail. We propose a new theory that captures both domain-averaged and local eddy heat flux behavior across all simulations. The theory uses scaling arguments for the zonal momentum balance and accounts for kinematic effects on eddy mixing. The scaling depends on a non-dimensional thermal Rossby number and implicitly accounts for the role of nonlinear eddy momentum fluxes across jets. We demonstrate the skill of the proposed scaling in explaining global near-surface temperature distributions across a wide range of climate states in a simple energy balance model. Implications for general circulation theory and extending the scalings to moist atmospheres are discussed.
Read the full abstract

Machine Learning

When Retain Constraints Conflict: Mitigating Forget-Retain Interference in Tabular Data

Machine unlearning aims to remove the influence of designated training data while preserving model utility, but its behavior on tabular data remains underexplored. This gap is important because tabular prediction is widely used in high-stakes domains and is increasingly adapted to language models through record serialization and schema-aware prompting. We identify a key challenge that distinguishes tabular unlearning from unlearning in free-form text or other modalities: schema-induced forget-retain overlap. In serialized tabular data, records share fixed column-name/value slots, similar attribute ranges, and common output spaces. Consequently, a forget row may have nearby retain rows that rely on the same high-signal attributes, causing retain preservation to oppose the update required for forgetting. Motivated by this failure mode, we propose Conflict-Aware Unlearning (CAU), a schema-aware approach that reduces forget-retain interference by relaxing preservation constraints on retained rows that most conflict with the forget set. Across sample-level and feature-level unlearning on clinical and non-medical tabular tasks, CAU more closely matches a retraining oracle while maintaining predictive utility and retain-region behavior. Our results show that reliable tabular LLM unlearning depends not only on the forgetting objective, but also on how retain constraints are constructed.
Read the full abstract

All authors Zijie Liu, Jinhao Duan, Bingqi Shang, Xinming An, Sijia Liu, Tianlong Chen

Computer Vision

Unified Multi-Layer Subspace Modeling for Cross-Domain OOD Detection

Out-of-Distribution (OOD) detection remains a fundamental challenge for neural networks, whose predictions can be overconfident on inputs that deviate from the training distribution. Most post-hoc OOD detection methods derive scores from a single representation level (eg., logits or penultimate features) or combine multiple layers via depth selection or OOD-calibrated weighting. However, because OOD shifts are diverse, the most informative representation level can vary strongly across OOD types and domains, making fixed-layer choices and OOD-calibrated aggregation brittle. In this paper, we propose PRISM (Projected Representation with Intermediate-layer Subspace Modeling), a model-agnostic post-hoc OOD detection method that models a unified multi-layer feature representation rather than aggregating independently scored layers. PRISM fuses intermediate and deep features into a single hierarchical embedding, estimates an in-distribution (ID) principal subspace, and then combines two complementary signals: (i) a class-conditional Mahalanobis distance in the projected subspace and (ii) the residual energy orthogonal to the learned manifold. This simple design avoids OOD-tuned layer weighting while capturing both in-subspace semantic deviations and off-subspace anomalies. Across diverse benchmarks spanning natural images, medical imaging, and industrial visual inspection, PRISM achieves consistent state-of-the-art cross-domain OOD detection performance with a single default configuration across all evaluated domains and architectures. We further show that PRISM incurs minimal inference overhead, making it practical for real-world deployment.
Read the full abstract

Computers and Society

GreenPassport: Request-Level Carbon Accounting for Cross-Border AI Inference

AI inference often crosses regional boundaries as prompts travel to remote data centers and generated tokens return to users. Regional averages cannot represent the resulting differences in serving hardware, electricity, and network delivery. Request-level accounting needs a common boundary for the service, serving site, route, local comparator, uncertainty, and data provenance. GreenPassport Carbon Accounting (GPCA) associates these inputs with each request. It estimates serving and route carbon, then selects a reporting level from the available documentation. Our public-data implementation covers data-center instances, accelerators, model families, electricity mixes, routes, and cloud-region carbon intensity. Against six accounting baselines and four energy-prediction baselines, GPCA reduced median absolute percentage error by 56.3% and median absolute error by 15.5% relative to EcoLogits under the aligned accelerator-energy boundary. It produced zero rule overstatement in the deterministic conformance tests. In the buyer case, the clean-electricity CN-West scenario produced 0.0148 gCO2e per request, 88% below the local service at 0.1220 gCO2e per request.
Read the full abstract

Cryptography and Security

AURA-Eval: Evaluation Framework for Acting Under Risk Awareness in LLM Agent Trajectories

LLM agents are increasingly taking actions with real consequences, yet safety evaluations often reduce their behavior to a final safe/unsafe outcome, obscuring whether the agent recognized the risk and how it responded. We introduce AURA-Eval, a framework for evaluating whether agents recognize contextual risk and adapt their behavior accordingly. AURA-Eval truncates tool-use trajectories at safety-critical decision points and asks the evaluated model to generate the next thought and action. It creates controlled scenario variations along six risk mechanisms: harm intensity, scale, target susceptibility, system dependency, oversight, and reversibility. It also creates paired scenarios that preserve the underlying goal but differ in whether safe fulfillment is available, allowing unsafe compliance and over-refusal to be evaluated together. An LLM-judge panel, validated against human annotations, evaluates each continuation along three axes: risk detection from the reasoning trace, and action type and safety from the action. From 157 R-Judge seed trajectories, we construct 1,249 evaluation items and evaluate 20 frontier, open-weight, and safety-tuned models. Across all 20 models, unsafe-action rates are higher when no safe path for request fulfillment is available than when one is; for frontier models, the rates are 27-40% versus 5-7%. When a safe path exists, frontier models rarely refuse; when none exists, they most often propose an alternative. Open-weight models, by contrast, more often execute the request than propose alternatives and act unsafely on 58-83% of items. Controlled risk-mechanism analyses further show particularly high unsafe rates under changes to oversight and scale; removing pre-execution oversight flips 52.3% of cases that were safe on the corresponding original scenario to unsafe.
Read the full abstract

All authors Ruoxi Shang, Christina-Maria Androna, Orfeas Menis Mastromichalakis, Yu Feng, Aniruddhan Ramesh, Rico Angell, Shang Hong Sim, Chrysoula Zerva, Emmanouil Koukoumidis

Combinatorics

A Uniform Product-Difference Theorem for Dense Subsets of $\mathbb Z^2$

We establish a uniform product-difference theorem for dense subsets of $\mathbb Z^2$, which gives an affirmative answer to Problem~2 of Fish and, as consequences, to both parts of his Problem~1. More precisely, we prove that for every $\delta>0$ there exists an integer $K(\delta)\geq 1$ such that every set $E\subseteq\mathbb Z^2$ with upper Banach density $d^\star(E)\geq\delta$ satisfies \[ K(\delta)\mathbb Z \subseteq \{ab:(a,b)\in E-E\}. \] As consequences, we obtain affirmative answers to both parts of Fish's Problem~1: for positive-density sets $E_1,E_2\subseteq\mathbb Z$ and $E\subseteq\mathbb Z^2$, respectively, the sets \[ (E_1-E_1)^2-(E_2-E_2)^2 \quad\text{and}\quad \{x^2-y^2:(x,y)\in E-E\} \] contain nontrivial ideals of $\mathbb Z$, with generators depending only on the corresponding density thresholds. In particular, the latter result also settles a conjecture of Davies concerning differences of the indefinite quadratic form $x^2-y^2$ in dense subsets of $\mathbb Z^2$.
Read the full abstract

Hardware Architecture

BLINK: Batch Normalization-based Integrity Checkpoints for In-Situ Detection and Mitigation of Diverse Weight Corruptions in DNN Accelerators

In safety-critical deployments, AI hardware must remain reliable against a broad spectrum of threats such as aging, soft errors, hard faults, and adversarial attacks (e.g. progressive bit flip attack (PBFA)). All of these corrupt stored weights while the chip keeps producing confident but inaccurate predictions. Detecting and mitigating such weight perturbations is crucial for safety-critical platforms. To that end, we propose BLINK, an on-chip batch normalization (BN)-based on-the-fly detection and mitigation approach, which is based on continual sensing of the shift in the activation statistics, and targets a wide variety of weight corruptions (random and localized faults as well as adversarial bit flips). BLINK operates in two phases: (1) off-line pre-characterization of the relationship of the activation shifts with inference accuracy drop, and (2) on-chip runtime detection and mitigation of weight corruptions. Upon detection, the flagged layer is re-centered to bring it closer to its stored clean reference within the same forward pass. BLINK is fully autonomous, eliminating the need for host communication, operation halts, or access to fine-tuning data. If the residual shift after mitigation indicates that accuracy has fallen below a user-set floor, a held-out watcher aborts the inference. Evaluated on ResNet-20/50 and MobileNetV2 for CIFAR-10/100, BLINK detects harmful corruptions with >99% precision across all fault types. Further, it recovers accuracy from 10% to 85.88% under 0.5% random bit flips (Resnet-50/CIFAR-10), up to 84% for localized faults (MobileNetV2/CIFAR-10), and from random-guess accuracy to 80%-83% under PBFA (ResNet-20/CIFAR-10). Hardware overhead estimates indicate that BLINK incurs negligible costs, with less than a 2% increase in latency and only a 0.53% increase in computation overhead.
Read the full abstract

Software Engineering

Shortcutting the Fix: Identifying and Categorizing Agentic Exploits in Software Engineering Benchmarks

High software engineering benchmark scores often mask exploitative shortcuts, such as retrieving upstream solutions, exploiting local Git histories, or recalling memorized code. We formalize this phenomenon as agentic shortcutting and audit three state-of-the-art open LLMs on SWE-bench Multilingual and DeepSWE using trajectory-level LLM judges. Standard evaluations trigger shortcutting attempts in 60.7%-90.1% and 55.5%-89.1% of trajectories, respectively, with up to 72.1% of tasks resolved with the aid of a successful exploit; disabling network access fails to curb these high attempt rates. Prompting agents to construct solutions independently suppresses attempts to 9.4%-14.2% and 0.9%-10.3%, respectively, while virtually eliminating shortcut-dependent resolutions and preserving core task performance. Our findings demonstrate the necessity of exploit-aware evaluation frameworks to measure repository-level problem solving.
Read the full abstract

All authors Nikolai Ludwig, Wasi Uddin Ahmad, Somshubra Majumdar, Boris Ginsburg

Machine Learning

DrugReason: Dynamic Multi-View Reasoning over Knowledge Graph and Language Evidence for Drug Repurposing

Drug repurposing aims to identify new therapeutic uses for existing compounds and, compared with de novo drug discovery, offers a faster and more cost-effective path to clinical translation. However, the space of candidate drug-disease pairs is enormous and their underlying relationships often depend on complex multi-hop biological mechanisms, making it difficult to reliably predict which pairs represent true therapeutic relationships. Existing approaches tackle this from two directions: knowledge graph-based methods organize curated biomedical evidence into structured relational networks for grounded multi-hop reasoning, while LLM-based methods leverage pretrained knowledge to generate flexible mechanistic rationales. Yet neither is sufficient alone - KGs are confined to observed graph structure while LLMs lack factual grounding and risk hallucination. To address this gap, we propose DrugReason, a multi-view reasoning framework that integrates grounded KG reasoning with LLM-generated mechanistic inference for drug repurposing. DrugReason adaptively routes diverse reasoning paths to specialized experts conditioned on the query context, while a cross-expert distillation objective enables knowledge sharing without sacrificing expert specialization. Experiments on PharmaDB, DDInter, and DrugBank show that DrugReason improves average performance over strong single-view reasoning baselines and achieves competitive or superior results compared with graph-based alternatives, while providing interpretable routing-based predictions.
Read the full abstract

All authors Zijie Liu, Hongxuan Li, Zhen Tan, Jinhao Duan, Baixiang Huang, Zunpeng Liu, Kai Shu, Tianlong Chen

Combinatorics

Positive Square Energy of Graphs with Minimum Degree at Least Two

Let $s^+(G)$ denote the sum of the squares of the positive adjacency eigenvalues of a graph $G$. The square-energy conjecture of Elphick, Farber, Goldberg, and Wocjan, proved by Liu, Tang, and Zhang, gives a lower bound of $n-1$ for any connected graph of order $n$. We strengthen this bound to $s^+(G)\ge n$ for every connected graph $G$ of order $n$ with minimum degree at least two, unless $G$ is a cycle.
Read the full abstract

Space Physics

AETHER-P3 Nowcast v1.0: Model Description, Training-Data Construction, and Validation Technical report prepared in support of CCMC onboarding

AETHER-P\textsuperscript{3} Nowcast v1.0 is a machine-learning-based global thermospheric neutral-density model developed for low-Earth-orbit applications and prepared for onboarding to NASA's Community Coordinated Modeling Center (CCMC). The model provides pointwise neutral-density estimates together with predictive uncertainty using a deep evidential regression framework driven by causal solar, solar-wind, geomagnetic, spatial, temporal, and empirical-model inputs. This report documents the released model configuration, training-data construction, software traceability, output products, validation strategy, benchmark performance, and known limitations. The training archive combines accelerometer- and mission-derived density observations from CHAMP, GRACE-A, GOCE, Swarm-C, and GRACE-FO spanning 2000–2023. More than 40 million eligible 30-s observations are available, but the archive is strongly dominated by consecutive quiet-time measurements. To preserve coverage of physically important regimes, the final 1.67-million-sample training set is constructed using deterministic regime-aware sampling that progressively subsamples quiet conditions while retaining all available extreme-condition observations. Validation uses temporally disjoint chronological blocks with exclusion guards to reduce information leakage. Evaluation across quiet, moderate, and extreme conditions shows competitive performance relative to HASDM, JB2008, NRLMSISE-00, and available WAM-IPE cases, while also identifying limitations associated with sparse training coverage, mission-dependent density products, and condition-dependent uncertainty calibration. The report provides a reproducible technical description of the AETHER-P\textsuperscript{3} Nowcast v1.0 research release and its current CCMC onboarding configuration.
Read the full abstract

Atomic Physics

Observation of the liquid-gas transition in trapped ions

Precision metrology, quantum information, and quantum-controlled chemistry studies all have many implementations in trapped ion systems that require a precise understanding of the motional dynamics of ions. Between the well-studied Coulomb-crystal and gaseous regimes lies the intermediate liquid-like regime whose properties and boundaries have largely eluded experimental investigation. Here, we track the entire liquid-to-gas transition process by leveraging direct velocity measurements using our newly developed ion-trap with integrated velocity map imaging. We demonstrate that the colder the ensemble, the more viscous the ion motion becomes until radial localization emerges, which marks the onset of the liquid regime at a temperature warmer than commonly assigned. At even hotter temperatures, we observe an abrupt change in the heating rate power-law, which indicates the maximum density threshold. The enhanced sensitivity of direct velocity measurements reveals the full transition process from a fluid with short-range order to a weakly-interacting gas.
Read the full abstract

All authors Eliana Ruth Wallach, Yohay Halfon, Yosef Alkoby, Yair Rajmiel, Nevo Werner-Reiss, Ilan Kleinman, Yuval Shagam

Computational Complexity

Improved Multilayered PCPs and Hypergraph Vertex Cover

We present two elementary constructions of multilayered PCPs that improve upon prior constructions in two ways. Specifically, we give one construction of quasi-linear size, and another one with $2$-to-$2$ constraints. Using these constructions we obtain the following results for the hypergraph vertex cover problem: $\bullet$ For $k=3$, for all $\varepsilon>0$, approximating the minimum vertex cover of a given $3$-uniform hypergraph within factor $1+\sqrt{2}-\varepsilon$ is NP-hard. Previously, the best known result due to [Dinur, Guruswami, Khot, Regev, SICOMP 2005] achieved a factor of $2-\varepsilon$. $\bullet$ For $k\geq 4$, for all $\varepsilon>0$, approximating the minimum vertex cover of a given $k$-uniform hypergraph within factor $k-\varepsilon$ is NP-hard, which is tight. Previous works established this result assuming the Unique-Games Conjecture [Khot, Regev, JCSS 2008], and a weaker factor of $k-1-\varepsilon$ for standard NP-hardness [Dinur, Guruswami, Khot, Regev, SICOMP 2005]. $\bullet$ Assuming the Exponential Time Hypothesis, for all $k\geq 3$ and $\varepsilon>0$ there is $C>0$ such that no $2^{n/\log^C n}$-time algorithm approximates the minimum vertex cover in a $k$-uniform, $n$-vertex hypergraph within factor $k-1-\varepsilon$. The proofs were obtained using ChatGPT 5.6 Pro and subsequently rewritten by the communicators.
Read the full abstract