Peer Revenue Returns subject to peer review

Papers

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

All of arXiv

Tuesday 1 September

Number Theory

Sharp lower bounds for shifted moments of Dedekind zeta functions

Let $K_1,\cdots,K_r$ be fixed number fields, and let $L$ be the compositum of their Galois closures. Assuming GRH for $\zeta_L$, we prove a sharp lower bound for products of shifted Dedekind zeta functions on the critical line, for arbitrary fixed positive real exponents and uniformly for shifts of size at most $T/2$. The correlation factor is expressed as a product of Dedekind zeta functions of the fixed fields of double-coset stabilisers in $\textrm{Gal}(L/\mathbb{Q})$. Combined with the corresponding upper bound by the authors, determines the order of magnitude of these shifted moments for both Galois and non-Galois fields.
Read the full abstract

Computation and Language

Beyond Magnitude: Contrastive Routing for Modular Mixture-of-Experts

In current Mixture-of-Experts architectures, routing is performed based on representations dominated by structure shared across all tokens, limiting expert specialization. We show that contrasting each token against an Exponential Moving Average of the layer's hidden states, rather than routing on absolute magnitude, concentrates the routing signal onto a low-dimensional, highly separable subspace. Building on this, we propose the Contrastive Routing Mechanism (CoRM), which scores each expert by the gap between its affinity for the incoming token and its affinity for this shared reference state, interpreted through a distinct per-expert projection. The resulting experts have routing boundaries that align with linguistic structure significantly more than the Top-k baseline. Our experiments show that CoRM improves average zero-shot accuracy by +0.67 to +1.69 points (Top-1) and +1.38 to +1.77 points (Top-2) over standard Top-k MoE baselines on nine zero-shot reasoning benchmarks, at the minimal cost of 2.9% added parameters and 2.6% added FLOPs per token.
Read the full abstract

All authors Nikolaos Xiros, Dimitrios Damianos, Maria-Eleni Zoumpoulidi, Leon Voukoutis, Vassilis Katsouros, Georgios Paraskevopoulos

Optimization and Control

The Complexity of Approximating the Value in Revealing POMDPs with Long-Run Average Objectives

We study partially observable Markov decision processes (POMDPs) with long-run average objectives, where the payoff is defined as the limit inferior of the expected average rewards. We consider the computational problem of approximating the long-run average value of a POMDP. In general, the long-run average value of a POMDP is neither computable nor approximable. We therefore consider the subclass of revealing POMDPs. Informally, a POMDP is revealing when the controller observes the underlying state with positive probability at each stage of the process. Our main contributions are threefold. First, we illustrate the practical relevance of this class of POMDPs through an application in control and optimization. Second, we present an exponential-time algorithm for the value approximation problem. Third, we establish EXPTIME-hardness by a reduction from the problem of almost-sure safety in POMDPs. Together, these results show that the problem of approximating the long-run average value in revealing POMDPs is EXPTIME-complete.
Read the full abstract

Analysis of PDEs

Gaussian upper bounds for averaged heat semigroups on graphs

Characterizations of pointwise Gaussian upper bounds on graphs with possibly unbounded geometry in terms of localized functional inequalities contain errors depending on the vertex degree. We introduce a new space-time averaged form of the heat semigroup in terms of time-averaged $\ell^{q/(q-1)}-\ell^q$-estimates and obtain Gaussian upper bounds from large-scale Faber-Krahn inequalities which avoid such errors. A main analytic ingredient is an integrated version of Davies' method which yields off-diagonal estimates for this averaged quantity, and which tends to pointwise Gaussian bounds as time tends to infinity. Conversely, on large scales volume doubling and Gaussian bounds on this averaged heat semigroup norms imply relative Faber-Krahn inequalities for subsets of prescribed relative measure. Our proof is based on a lower bound for Dirichlet eigenvalues in terms of these averages. The Faber-Krahn dimension is scale-dependent, but converges to the doubling dimension for increasing radii. This gives a characterization of asymptotic Gaussian heat kernel behavior in terms of Faber-Krahn inequalities.
Read the full abstract

High Energy Physics (Theory)

Galilean Kalb-Ramond Field

We construct Galilean limits of the free Kalb–Ramond two-form gauge field and analyze their symmetries using two different procedures. An Inönü–Wigner contraction of the $D$-dimensional equations of motion results in two inequivalent electric and magnetic limits. In both cases the equations are invariant under the complete infinite-dimensional Galilean conformal algebra (GCA) only at $D=6$, the critical dimension of the relativistic theory. The second procedure is a null reduction of the $(D+1)$-dimensional theory which yields a local action for a four-field multiplet. In the zero-momentum sector, the resulting action is invariant under translations, rotations, Galilean boosts and an anisotropic dilatation with dynamical exponent $z$. However, in contrast to Galilean electrodynamics, it is not invariant under the special conformal generators of the GCA (and its higher modes). We attribute this to the vector residing at the bottom of the boost chain. If the parent theory is taken to be the six-dimensional conformal one, the resulting five-dimensional reduced action is invariant under the Schrödinger special conformal transformation with $z=2$ which is derived from the relativistic conformal group. In light-cone gauge, the non-zero null momentum sectors reduce to free Schrödinger fields. We obtain the gauge-fixed propagator, which is instantaneous, and compare the free-field correlators with the two-point functions allowed by the Ward identities of both.
Read the full abstract

Cryptography and Security

CRSF: Collusion-Resilient Privacy-Preserving Sensor Fusion with Byzantine-Robust Participation

Privacy-preserving sensor fusion enables an untrusted server to compute an aggregate result over distributed sensor measurements without learning either individual inputs or the final output. Recent garbled-circuit-based protocols provide an efficient realization of this functionality in a sensor–server–client architecture, but remain vulnerable to sensor–server collusion and Byzantine manipulation of sensor participation. These weaknesses can compromise honest-sensor privacy, incorrectly exclude honest sensors, and corrupt the computed fusion result, thereby undermining the security guarantees expected from the protocol. We present CRSF, a collusion-resilient sensor-fusion protocol that addresses these weaknesses while providing privacy, correctness with explicit abort, and liveness. CRSF introduces a Practical Byzantine Fault Tolerance (PBFT)-based agreement phase for sensor submissions and uses server-specific, status-dependent label release with threshold protection of circuit-input labels. This design prevents any Byzantine server from unilaterally manipulating sensor participation and prevents any admissible sensor-server coalition from obtaining enough secret material to compromise honest-sensor privacy. We implement CRSF and compare its online execution time with the most relevant state-of-the-art baseline. Our Google Cloud evaluation measures the total computation and communication cost of the online protocol under fault-free and representative faulty executions. Across a range of fault-tolerant fusion circuits and up to 261 sensors, CRSF demonstrates a highly practical trade-off between robust security and protocol performance.
Read the full abstract

All authors Chao Yin, Haihong Tian, Zheng Yang, Haibin Zhang, Fabio Massacci, Chenglu Jin

Software Engineering

Reliable LLM-Generated Programs for High-Energy Physics Experiments through Graph-Grounded Software Knowledge

Extracting physics information from modern particle-physics experiments requires multistage analyses implemented on top of large and highly interconnected software ecosystems. General-purpose large language models (LLMs) often produce unreliable programs for such tasks because a user request alone rarely specifies the required APIs, dependencies, and usage conventions. We organize these software relations before generation and retrieve task-relevant knowledge at inference time. Using the open-source ROOT framework as a representative and reproducible testbed, we evaluate a complete grounding system that combines hybrid retrieval over a heterogeneous software knowledge graph, skill-selected workflow examples, and execution-guided repair. On a benchmark of 275 ROOT tasks, grounding improves first-attempt execution from 58.5% to 76.0% under Claude Code orchestration and from 51.3% to 64.0% under standalone orchestration. Final success increases from 90.5% to 96.0% and from 78.9% to 90.9%, respectively, while the average generation cost per successful task increases by only 1.3% and 3.2%. The gains persist under a strong coding agent, indicating that explicit software knowledge remains valuable even when agentic scaffolding is already in place. Because the method captures software relations common to large codebases rather than facts specific to ROOT or a particular model, it should transfer to other experiment frameworks and proprietary software, especially where documentation is sparse or internal dependencies are complex.
Read the full abstract

All authors Yue Sun, Tong Liu, Yipu Liao, Jingde Chen, Ke Li

Differential Geometry

New extremal Kähler metrics on projective bundles

Consider a holomorphic vector bundle $E$ over a compact complex curve $C$ which decomposes as a sum of stable vector bundles. For the projectivization $\mathbb{P}(E)$, we prove that the existence of a compatible extremal almost Kähler (aK) metric of involutive type in the sense of Lejmi is equivalent to the existence of a Calabi extremal Kähler metric. This result rests on the Yau–Tian–Donaldson correspondence in terms of the moment polytope $\Delta$ for $\mathbb{P}(E)$, proved by the author and Yin in a previous work. The main advantage is that compatible extremal aK metrics of involutive type are solutions to a second-order linear PDE, rather than a fourth-order nonlinear PDE for Calabi's extremal Kähler metrics. As an application, we prove that when $E$ has rank $4$ and $C$ is an elliptic curve or the projective line, $\mathbb{P}(E)$ is a Calabi dream manifold, i.e. admits an extremal Kähler metric in every Kähler class.
Read the full abstract

Analysis of PDEs

Ideal MHD below the classical well-posedness threshold

We establish local existence and uniqueness of solutions for the ideal incompressible magnetohydrodynamics system posed on $[0,T]\times\mathbb{R}^n$ with $n\ge2$ and initial conditions given by a nonzero constant magnetic field $\mathbf{B}_0$ and arbitrary divergence-free velocity $v_0\in H^s$, in the range $(n+1)/2<s\le n/2+1$. The proof uses a Lagrangian wave–Hodge reformulation and exploits an Alfvén null–structure hidden in the pressure forcing. This work builds on Zhang \cite{Zhang2026} and provides a bridge from the Lagrangian theory to the Eulerian Cauchy problem. In particular, the constructed solutions are induced by a bi-Lipschitz and volume-preserving map.
Read the full abstract

Computer Vision

IT-TextFusion: Iterative Text-Image Interaction with Text-Guided Residual Refinement for Degradation-Aware Image Fusion

Text-guided image fusion has recently emerged as an effective paradigm for integrating multi-modal information while enabling flexible and task-oriented fusion control. However, existing text-guided fusion methods often rely on shallow semantic-visual interaction and limited attention mechanisms, which restrict their ability to robustly handle complex degradations and fully exploit textual guidance. In this paper, we propose an iterative text-guided image fusion framework that incorporates text-conditioned feature interaction across multiple fusion and refinement stages. The proposed method integrates deepest-level Cross-Attention, multi-scale Cross-Gate Fusion, and stage-specific text-conditioned modulation, allowing the global text embedding to condition hierarchical feature fusion and residual refinement. By repeatedly injecting the pooled text embedding across hierarchical decoder and refinement stages, the proposed framework provides degradation-aware global semantic conditioning while preserving complementary information from the visible and infrared modalities. Experiments on several benchmark datasets show that the proposed method improves several information-preservation and perceptual-quality metrics, while exhibiting metric-dependent trade-offs on some datasets.
Read the full abstract

All authors Siyang Liu, Peiyi Zhou, Tianle Jin, Rongrong Bian, Zheke Jin, Mengze Gao

Machine Learning

Subliminal Learning as Trait-Direction Drift: A Mechanism and Targeted Control under SFT Distillation

Beyond intended capabilities, model distillation can transfer hidden traits from a teacher. A teacher biased by a system prompt can generate semantically clean training data, such as numeric sequences, that still causes a downstream student to inherit the hidden preference, a phenomenon known as subliminal learning. Prior work has identified several parts of this process. How the signal builds up during training and produces behavioral transfer remains unclear, making targeted mitigation difficult. We propose and validate trait-direction drift as a mechanism for subliminal learning: biased generation creates measurable preference gaps in teacher data, and student-recognizable gaps induce trait-aligned updates during supervised fine-tuning that accumulate into behavioral transfer. Guided by this mechanism, we propose probe-space corridor regularization, a targeted defense that constrains drift along a calibrated trait direction during distillation. The method substantially reduces hidden-trait transfer, preserving task performance: for example, it lowers malicious-response transfer from 29.55% to 6.45% with low main-task accuracy cost, and consistently suppresses animal-preference transfer across the main Qwen setting. The preference-gap, training-trajectory, and intervention evidence links subliminal learning to trait-direction drift and motivates corridor regularization as a targeted control during distillation.
Read the full abstract

All authors Zhixuan Liu, Zhichen Dong, Yuyu Fan, Xiangtian Li, Chao Yang

Machine Learning

Modelpedia: A Catalog of Model Findings for the Meta-Science of AI

Scientific knowledge about AI models is produced faster than the community can organize it. Every few months a new foundation model reshapes the field and hundreds of papers, blogs, and technical reports document how each behaves or fails. Yet, these findings remain scattered and effectively unretrievable. To address this gap we present Modelpedia, an automated, LLM-assisted framework that extracts findings about models from published papers, links it to the model, dataset, method, and concept it concerns, and aggregates the result into a searchable public catalog. Applying the prototype to accepted ICLR 2024 and 2025 papers, we extract over a thousand findings and, treating the catalog itself as an object of study, run a meta-analysis of how the community investigates models. Now, we invite the community to explore, contribute to, and build on the open catalog, and to help establish model findings as a shared foundation for the meta-science of AI.
Read the full abstract

All authors Franciszek Bernat, Dawid Płudowski, Michał Jan Włodarczyk, Luca Longo, Jianlong Zhou, Andreas Holzinger, Riccardo Guidotti, Wojciech Samek, Przemysław Biecek

Robotics

Adaptive Depth-Map-Guided Bundle Adjustment for Correspondence-Free Multi-View Point Cloud Registration

Robotic processing of irregular steel scrap requires dense 3-D measurement to replace manual visual assessment in hazardous cutting workcells. The reconstructed map is used to estimate piece dimensions, boundary geometry, feasible preheating and cutting regions, and collision-aware torch paths. The reconstruction errors therefore propagate directly to downstream measurement and planning. Existing multi-view registration methods commonly rely on feature extraction and data association to establish correspondences between views. In workcells with smooth metallic surfaces, repeated structures, occlusions, and partial overlaps, however, wrong correspondences may be established, leading to inaccurate pose estimation and distorted reconstruction. This paper presents an adaptive layered depth-map-guided bundle adjustment framework for correspondence-free multi-view point cloud registration. The scene is represented by a global 2.5-D grid, where each cell can adaptively maintain multiple depth hypotheses. Raw depth observations are directly projected into the global map to form depth constraints without explicit feature correspondences. At grid cells where multiple surfaces produce conflicting depths, a softmax-based layer assignment links each observation to compatible depth hypotheses. The resulting nonlinear least-squares formulation jointly refines sensor poses and the layered depth map, with correspondences implicitly induced by the depth-map representation and projection model. Experiments on self-collected industrial datasets show that the proposed method achieves consistently competitive reconstruction accuracy while maintaining robustness and low computational cost in challenging industrial scenarios. We release the open-source code implementation at: https://github.com/YiranZhou-Robotics/ADM-BA.git
Read the full abstract

All authors Yiran Zhou, Yingyu Wang, Shoudong Huang, Liang Zhao

Materials Science

Pressure-regulated mechanochemistry at lithium metal-sulfide electrolyte interfaces

Stack pressure is commonly treated as a means of maintaining physical contact in all-solid-state lithium-metal batteries, but it can also alter the chemistry of reactive solid-solid interfaces. Here, using pressure-aware, charge-resolved machine-learning molecular dynamics validated against DFT, we determine how pressure magnitude and loading geometry regulate interphase formation at Li||Li6PS5Cl interfaces. The response is nonmonotonic: compression at 1 kbar accelerates PS4 decomposition and Li2S-like ordering, whereas 10-100 kbar compression restricts structural rearrangement and long-range crystallization. Charge-resolved dynamics further identify sulfur-centered, lithium-rich early-interphase environments associated with subsequent Li2S-like ordering. Uniaxial loading accelerates interfacial reaction relative to isostatic loading at the same nominal pressure. Pressure also changes void closure and dead-lithium spreading in a defect-location-dependent manner. These results establish applied pressure as a mechanochemical process variable coupling interphase chemistry, ion transport and defect evolution, providing a mechanistic framework for interpreting pressure effects in sulfide solid-state batteries.
Read the full abstract

All authors Kunik Jang, Jaehwan Choi, Jang Wook Choi, Yousung Jung

Software Engineering

Fine-Tuning Large Language Models to Classify Pull Request-Issue Alignments: Going Beyond Prompting

Context: Accurate alignment between pull requests (PRs) and corresponding issues is crucial for efficient software development and maintaining code quality, as misalignments can reduce traceability, hinder defect localization, and decrease maintainability. Objective: This study aims to improve automated PR-issue alignment classification by leveraging fine-tuned large language models (LLMs) across multiple alignment categories, and conducts interpretability analysis to investigate the effects of PR-issue fields on the predictions of fine-tuned LLMs. Method: Our methodology consists of dataset preparation, LLM fine-tuning, and interpretability analysis. We first extended an existing dataset and applied data augmentation to address class imbalance. GPT-4o was then fine-tuned via instruction tuning, and open-source LLMs including CodeLlama-7B, CodeQwen1.5-7B, StableCode-3B, CodeGemma-7B, and Deepseek-Coder-6.7B were fine-tuned using classification-specific heads. Interpretability analysis using Shapley Additive Explanations (SHAP) was conducted to examine the influence of PR-issue fields on predictions for the best-performing open-source LLM. Results: Fine-tuned LLMs outperformed baseline models, achieving average improvements of 6.15% in accuracy and F1-micro, 14.69% in F1-macro, and 6.15% in recall. CodeLlama-7B emerged as the best-performing fine-tuned LLM overall, while interpretability analysis revealed that code diffs together with issue body and PR body contents exert the greatest influence on predictions. Conclusions: Fine-tuning substantially enhances PR-issue alignment classification, improving both accuracy and efficiency. Interpretability analysis provides actionable insights into the dataset features driving alignment decisions, deepening understanding of how LLMs reason over software artifacts.
Read the full abstract

Combinatorics

The real-rootedness of the toric $g$-contribution polynomials

Recently, Ehrenborg, Hetyei and Readdy expressed the toric $g$-polynomial of a simple polytope as a linear combination of a family of polynomials, called $g$-contribution polynomials, with coefficients given by the entries of its gamma-vector. They conjectured that these toric $g$-contribution polynomials are real-rooted. This paper proves this conjecture.
Read the full abstract

Medical Physics

PyDoseRT Photon: Physics-Guided Pencil-Beam Dose Calculation with Neural Priors and Residual Correction for CT and MRI

We present a hybrid, physics-based analytical pencil-beam (PB) dose engine, augmented by two small frozen neural physics priors, followed by a 3-D convolutional residual-correction network (U-Net). We address the DoseRAD2026 (https://doserad2026.grand-challenge.org/) photon dose-prediction task with PyDoseRT Photon, a GPU PB engine implemented in PyTorch and corrected toward Monte Carlo (MC) accuracy in three learned stages of decreasing physical specificity. The engine reproduces the challenge's head-less MC source exactly where it can and models the patient with a beam-quality-indexed pencil kernel evaluated at the field's fluence-weighted radiological depth, and TERMA source scaling at the interaction site. Two tiny neural priors are trained through the frozen engine and then frozen themselves: a 39k-parameter 2-D fluence correction and a 48-parameter lateral heterogeneity correction mixing mass-conserving Gaussian redistribution operators. A compact 3-D U-Net (1.36M parameters) with a sequential refinement branch then predicts, per control point (CP) in the beam's-eye-view (BEV) frame, a bounded multiplicative gain and additive residual from seven channels. All learned stages are zero-initialized, so training starts from the analytical solution. For MRI, an nnU-Net regression model synthesizes a CT that enters the identical pipeline, with consecutively, the same trained corrector as the CT track. Design choices were driven by the challenge ranking, in which runtime carries double weight. The submitted method evaluated on our local CT and MR validation dataset achieved CP MAE 0.0086 and 0.0098, IDD distance 0.0011 and 0.0013, plan MAE 0.0024 and 0.0047, gamma pass rate (1%/1mm) 99.12% and 96.95%, with runtimes of 46s and 49s, respectively.
Read the full abstract

All authors Attila Simkó, Lukas Zimmermann, Hermann Fuchs, Gerd Heilemann

Hardware Architecture

Hardware Acceleration of Block-Diffusion LLM for Edge Devices

Single-stream (batch-one) edge inference cannot amortize weight traffic across requests. Full-attention diffusion LLMs recompute the entire sequence at every step; native block diffusion makes completed blocks immutable and exactly cacheable, yet refinement still streams prefix KV and FFN weights. We co-design WIFiV-LPDDR, a wide-I/O LPDDR system for precision-tagged reads, BRQ-KV for a canonical low-rank-plus-INT8-residual prefix with query-dependent per-entry precision, and DAT-FFN for drift-mapped canonical replacement, adjacent-stage-corrected low-bit delta, or cached-state carry while keeping live activations unquantized. Both map to an input-stationary mixed-precision systolic array. For the evaluated 1.5B/7B models on modeled Jetson-class platforms, the full stack provides arithmetic-mean energy-reduction factors of 3.79x/3.96x and arithmetic-mean latency speedups of 2.88x/4.44x at the reported DAT-FFN settings; every corresponding compressed model-benchmark score drops by less than one absolute percentage point from its baseline.
Read the full abstract

All authors Wei-Hsing Huang, Kiseok Lee, Ming-Yen Lee, Weiyu Sun, Cheng-Jhih Shih, Gayatri Tanksali, Arpit Khandelwal, Pin-Jun Chen, Yingyan Celine Lin, Shimeng Yu

High Energy Astrophysical Phenomena

The thermal emission for knot A of radio galaxy M87 by Chandra

The jet may compress an interstellar medium and soft X ray outside the jet may be absorbed by a compressed interstellar medium and thermal emission heated by a shock is expected inside the jet. Thermal emission was found for HST-1 and knot A of M87 in X ray energy spectra analysis by Chandra (Osone, 2017). The X ray image of the surrounding of the jet of M87 was studied by Chandra (Osone 2024). I find a soft X ray dip in a south region outside knot A. Therefore, by taking only a north region outside knot A as a background, where there is no soft X ray dip, I analyze X ray energy spectra for knot A by Chandra and confirm thermal emission with a high significance.
Read the full abstract

Distributed Computing

Update for Decisions, Not Freshness: Goal-Oriented Status Updating and Selective Offloading at the Network Edge

In an edge–cloud collaborative edge-computing environment, an edge node (EN) must decide whether each user task should be executed locally, forwarded to a remote service (or cloud) node (SN), or rejected. The EN observes its local state directly but receives the SN state only through an intermittently refreshed cache. Status updating and task control therefore form an asynchronous closed loop under partial observability. Freshness-driven schemes, including those based on Age of Information (AoI), do not directly value an update by its effect on subsequent task decisions. We propose CoSMO (Co-design of Semantic-state Management and Offloading), a cooperative event-driven reinforcement learning (RL) framework that coordinates semantic status management and selective offloading through realized task utility. CoSMO learns a compact representation of the heterogeneous SN service state. At the SN, a recurrent semi-Markov double deep Q-network (Double DQN) agent jointly selects send/no-send and the next decision interval. At the EN, a task-terminal off-policy value-learning agent makes hierarchical gate–route decisions from local observations and stale remote semantics. The agents maintain separate observations and value targets but share the same realized task-utility stream, without centralized execution. Across the evaluated workload families, CoSMO's reported relative improvement in on-time completion rate over the best-performing competing method averages 18.6%–21.2%. For capacity-aware decision accuracy across the three strict-overload points, the corresponding reported gains average 17.6%–$17.9%.
Read the full abstract

All authors Jianpeng Qi, Qiyang Zhang, Chao Liu, Jing Sun, Yimei Liu, Yanwei Yu, Yingjie Wang, Wei Ni

Computation and Language

StateSwap: Probing Support-Elimination Hidden States in Multiple-Choice Questions

Large language models often answer the same multiple-choice question inconsistently when it is posed under support-oriented and elimination-oriented framings. We investigate whether these discrepancies arise from different internal representations induced by the two framings. We introduce a dual-framing protocol with minimally varied prompts that use either support- or elimination-oriented framing while keeping the evaluation target fixed. To probe the internal computation, we append an untrained special token, [STATE], and treat its residual-stream activation as an intervention interface. Across both models, the two framings induce separable [STATE] activations concentrated in intermediate layers. Swapping these activations between paired prompts systematically changes predictions and improves cross-framing agreement, providing intervention-based evidence that the activations are behaviorally relevant. Beyond instance-level substitution, mean-difference steering directions derived from the dual-framing contrast exhibit more bounded layer-wise responses than matched contrastive activation addition directions under the evaluated protocol.
Read the full abstract

All authors Chao Gao, Haijiang Liu, Qiyuan Li, Caicai Guo, Frank van Harmelen, Jinguang Gu

Materials Science

Trap Dynamics and Conductivity Changes in AlGaN/GaN Heterostructures Under Ultrasonic Loading

AlGaN/GaN heterostructures are critical for high-power electronic devices but suffer from electron trapping at defects, limiting reliability. We investigate how ultrasonic vibrations affect electron transport in MOCVD-grown AlGaN/GaN heterostructures by combining temperature-dependent Hall effect and high-resolution X-ray diffraction measurements. Here we demonstrate that ultrasonic loading induces persistent acoustoconductivity and lattice parameter changes, attributed to acoustically driven rearrangement of metastable DX centers. This leads to increased carrier concentration and decreased mobility, reflecting defect state modulation by acoustic strain. These findings provide new insights into trap dynamics under dynamic deformation and suggest ultrasonic treatment as a potential approach to mitigate trapping effects, thereby enhancing the performance and reliability of GaN-based devices.
Read the full abstract

All authors Vladyslav Kaliuzhnyi, Mykola Tymochko, Oleksandr Gudymenko, Oleg Olikh, Alexander Belyaev

Metric Geometry

A threshold phenomenon for embeddings of Euclidean snowflakes and impossibility of dimension reduction

Fix $0<\theta\leqslant 1$. We prove that if $1\leqslant p \leqslant 2/\theta$, then the $\theta$-snowflake of $\ell_2^k$, namely, $\mathbb{R}^k$ equipped with the metric $((x,y)\in \mathbb{R}^k\times \mathbb{R}^k)\mapsto \|x-y\|_2^\theta$, embeds with distortion $O(1)$ into $\ell_p^m$ for some integer $m\lesssim_{p,\theta}k$, which is optimal as $k\to \infty$, as seen by comparing dimensions. However, for $p$ larger than the sharp threshold $2/\theta$ the following change in behavior occurs: If a $(1/\sqrt{k})$-dense subset of the Euclidean sphere $S^{k-1}$ embeds into $\ell_p^m$ with distortion $O(1)$, then necessarily $m\gtrsim_{p,\theta}( k/\log k)^{p\theta/2}$, which grows super-linearly in $k$ as $p\theta/2>1$, and this dimension bound is optimal as $k\to \infty$ up to lower order factors. We deduce from this statement that if $2<p<\infty$, then there exist arbitrarily large $n$-point subsets of $\ell_p$ with the property that if they embed with distortion $O(1)$ into $\ell_p^m$, then necessarily $m\gtrsim_p ((\log n)/(\log\log n)^2)^{p/2}$, thus demonstrating that the statement of the Johnson–Lindenstrauss dimension reduction lemma fails to hold for $\ell_p$
Read the full abstract

Statistical Mechanics

Statistical Language Competition Model with Dynamic Edge Weighting on a Random Network

This paper presents a computational study of language competition dynamics on Erdős–Rényi random networks, extending the foundational Abrams–Strogatz model through two novel contributions: (i) a dynamic edge-weighting mechanism that reinforces social ties between co-minority speakers by an additive increment $\Delta$, and (ii) a probabilistic agent-based framework governing language switching via a weighted majority rule. Phase boundaries separating the dominance and coexistence regimes are identified across a two-dimensional parameter space $(p, \Delta)$, where $p$ denotes the network connectivity probability. We further characterise anomalous persistence zones within predicted dominance regions, attributing them to the formation of isolated minority speaker clusters. Scaling study across network sizes $N \in \{50, 100, 250, 500, 1000\}$ reveal that average cluster size decreases with $N$ and that phase boundaries diffuse with increasing stochastic noise. Finally, we discuss extensions to a tripartite bilingual model and heterogeneous prestige/volatility to more faithfully capture real sociolinguistic contact scenarios.
Read the full abstract

Cryptography and Security

JENGA: Exploiting Counter-Based RowHammer Countermeasures to Break Real-Time Predictability

Safety-critical real-time systems must satisfy multiple dependability requirements, notably time predictability and security. In such systems, tasks must complete within bounded and known execution times, typically characterised through Worst-Case Execution Time (WCET) analysis. At the same time, DRAM-based platforms are increasingly sensitive to the RowHammer read-disturbance security vulnerability, which has motivated the development of numerous hardware and software countermeasures in both academia and industry. However, the impact of these defences is generally evaluated in terms of average-case performance, a metric that is insufficient for safetycritical real-time systems, where worst-case behaviour is the primary concern. In this paper, we study the impact of RowHammer countermeasures based on hardware counters on the timing behaviour of real-time systems. We use a Per-Row-Activation-Counter (PRAC) countermeasure as a case study, standardised for recent DDR5 memories, and show that it can introduce significant timing variations. Based on this observation, we introduce JENGA, an attack in which an attacker-controlled task manipulates the internal state of the RowHammer countermeasure mechanism to increase the execution time of a victim real-time task beyond its expected WCET. We implement JENGA in a gem5 and Ramulator 2.0 simulation environment and evaluate its impact on TACLeBench workloads. We show that such an attack can delay tasks up to 200% of their WCET, making the initial timesafety assumptions unsafe. To address this issue, we derive a safe analytical bound that accounts for mitigation-induced delays in WCET analysis for DRAM systems protected by hardware countermeasures, such as PRAC-N.
Read the full abstract

All authors Valentin Abgrall, Marcello Traiola, Ruben Salvador, Maria Méndez Real, Alessandro Palumbo, Angeliki Kritikakou

Materials Science

Text-guided flow matching enables sample-efficient crystal structure generation

Crystal generators can now propose periodic structures, but their control interfaces remain poorly matched to the mixed descriptors used in materials design. Text provides a compact way to combine composition, symmetry, prototype and property cues, yet it has not been clear whether such information can steer flow-based crystal generation. Here we introduce TFMat, a text-conditioned flow-matching framework that uses structured materials language as a semantic prior for a CrystalFlow generator. Across Perov-5, Carbon-24 and MP-20 crystal structure prediction benchmarks, TFMat improves one-candidate match rates over CrystalFlow and reaches a 92.04% MP-20 match rate with 20 candidates; in de novo generation, it improves element-count and density distribution alignment while retaining coarse property consistency in composition-selected outputs. These results position structured text as an inspectable control layer for translating human-readable materials intent into candidate crystals for downstream simulation and validation.
Read the full abstract

Networking

Lacan: Making Accountability in Anonymous Networks Real

Anonymity and accountability are essential properties for our everyday activity on the Internet. However, they appear contradictory, and their reconciliation remains far from reality. Existing approaches fall short in this regard, as they either rely on an on-path trustee, per-packet authorization, per-packet public-key cryptography, or per-session intervention by a central authority. We propose Lacan, a protocol that reconciles anonymity and accountability within a realistic design. In Lacan, a sender enjoys anonymity provided by on-path relays, as long as she complies with a contract established with the receiver. Upon a contract violation, the verifier, an off-path trustee on behalf of the receiver, links the malicious message to the sender's identity indirectly via the packet, path, and session, thereby reducing public-key operations from per-packet to per-session. This linkage remains robust even against malicious relays and receivers, grounded in our novel chain of successor proofs for accountable path reconstruction, together with traceable signatures, path validation, and key-committing encryption. We analyze the anonymity and accountability, implement the protocol, and evaluate the performance.
Read the full abstract

All authors Naoya Takada, Yutaro Yoshinaka, Kentaro Kita, Junji Takemasa, Yuki Koizumi, Toru Hasegawa

Signal Processing

OTFS Channel Estimation Utilizing Sparse Bayesian Generative Modelling

One of the key challenges of future wireless communication systems is ensuring reliability in high-speed mobile scenarios, where accurate recovery of channel state information (CSI) is essential. Many recent studies have concluded that orthogonal time-frequency space (OTFS) modulation is a promising technology for addressing this challenge. Additionally, machine learning (ML)-based methods have the potential to improve channel estimation performance by leveraging ambient information more effectively than classical estimation techniques. This paper particularly addresses channel estimation for OTFS by employing a compressive sensing (CS)-based sparse Bayesian generative model (SBGM), namely the recently introduced compressive sensing Gaussian mixture model (CSGMM). We show that our proposed approach yields significant improvement in normalized mean squared error (NMSE) over the next-best-performing baseline. We additionally provide insights into the theoretical potential of the model to optimally approximate complex channel distributions with arbitrary precision within the Doppler-delay (DD) domain. To summarize, this work establishes the OTFS-CSGMM framework as a promising solution for high mobility wireless channel estimation.
Read the full abstract

All authors Louis Anseaume, Benedikt Böck, Franz Weißer, Wolfgang Utschick

Computation and Language

Post-hoc Alignment of LLM-judges to Human Judgment Distribution

The LLM-as-a-judge (LLMaJ) framework offers a cost-effective and reproducible solution for automatic evaluation. However, current evaluation practices typically compare LLMaJ judgments against aggregated ground-truth labels, overlooking the valuable information contained in Human Label Variation (HLV). Inspired by an increasing line of work that proposes to leverage HLV, we systematically study LLMaJ performance on predicting both a single, aggregated ground truth hard-label and unaggregated soft-labels that represent Human Judgment Distributions (HJD). Our results across five diverse datasets reveal that while LLMs achieve near human-level performance at hard-label prediction on most tasks, they exhibit poor performance when predicting soft-labels. To address this limitation, we propose NAPHA (eNtropy-Aware Post-Hoc Alignment), a simple yet effective lightweight post-hoc alignment method that matches the LLM distribution to the HJD by first assigning an instance to a discrete entropy class and then routing it to specialized, trained alignment models. We find that NAPHA consistently improves soft-labels prediction across base LLM models and datasets, with particularly strong gains on high-entropy instances where capturing diverse human perspectives is most critical. We also show via oracle experiments that improving entropy class prediction can substantially enhance NAPHA's practical effectiveness.
Read the full abstract

All authors Sebastian Steindl, Nikos Voskarides, Alberto Gasparin, Diego Marcheggiani

Machine Learning

Let Confidence Change, Not the Prediction: Prediction-Preserving Repair for Post-hoc Calibration

Post-hoc calibration corrects reported confidence, yet a multiclass calibrator can also change the associated top-1 prediction. Accuracy captures only the net effect of these changes on correctness, not how often predictions change; the Top-1 Prediction Change Rate (TPCR) instead measures this frequency. We propose Calibrator-Output Repair for Top-1 Decision Preservation (CORD), the first post-fit adapter to impose exact prediction preservation by repairing the full calibrated probability vector. From the original and calibrated outputs alone, CORD determines the mass assigned to the original top-1. The calibrated conditional distribution allocates the remaining mass over the other classes, yielding a repaired vector whose own argmax recovers the original prediction. On the calibration split, CORD coordinates the repaired masses to retain the calibrated outputs' mean mass on original predictions whenever attainable. The adapter alters neither the fitted calibrator nor its direct output, fits no additional supervised map, and requires no user- or validation-tuned hyperparameter. Across CIFAR-10/100 and ImageNet-1K, CORD attains zero TPCR by construction and lowers mean ECE, NLL, and Brier relative to the corresponding direct outputs in every dataset; paired gains persist under distribution shift and across calibration-set sizes. CORD thus removes the preservation constraint from calibrator fitting and assigns exact recovery of the original decision to subsequent output repair. Our code is available at https://github.com/labhai/CORD.
Read the full abstract