Peer Revenue Equity in science

Papers

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

All of arXiv

Friday 28 August

Mesoscale and Nanoscale Physics

ScatterWorks: A Python package for building and solving scattering network models

Network models provide an efficient framework for studying non-interacting transport and wave-propagation phenomena in disordered and topological systems. We introduce ScatterWorks, an open-source Python package for building and solving network models. The package uses a compact network representation with composable transformation operations (tile, cut, relink, and union) to enable flexible construction of periodic and finite networks. Local scattering matrices on nodes can be assigned by node labels or explicit indexing, assembled into global scattering equations, and used to compute quasienergies and transport observables. The package supports sparse scattering matrices, allowing transport observables to be evaluated with Schur-based solvers. We demonstrate the package workflow on a symbolic Fabry-Perot interferometer and a numerical Chalker-Coddington model, recovering analytical expressions and the expected near-critical transport behavior of the quantum Hall transition. Plotting utilities with label-aware rendering are included to streamline debugging and reproducible setup for larger, custom network geometries.
Read the full abstract

Computation and Language

Blind Men and the Elephant: Probing the Epistemic Myopia of LLMs under Long-Tail Divergent Knowledge

Factual question answering (QA) typically assumes a single canonical answer, obscuring whether large language models (LLMs) retain divergent accounts of long-tail facts. To address this gap, we introduce ElephantBench, a closed-book knowledge probe comprising 1,094 questions generated through an auditable graph-based pipeline. The pipeline retrieves related documents from a low-exposure web corpus, identifies naturally occurring disagreements, and converts them into multi-account QA records. Each answer is verified against the originating documents and authoritative public web sources and is then reviewed by human annotators. Across 32 models, even the strongest model recovers both accounts on only 52.4% of questions, while on nearly all remaining questions it recalls one account but omits the other. Scaling model size and inference-time reasoning improve recall but do not eliminate this incompleteness. Corpus analysis further shows that exposure imbalance favors the dominant account, whereas greater minority-side exposure is associated with more complete recall. These findings establish ElephantBench as a reproducible knowledge probe for diagnosing epistemic myopia in parametric memory. More broadly, our graph-based benchmark construction pipeline provides an efficient and scalable way to turn long-tail corpora into source-traceable knowledge probes, supporting efforts to evaluate and advance the epistemic rigour of next-generation LLMs. Code is available at https://github.com/Tencent/ElephantBench.
Read the full abstract

All authors Zhuoshi Pan, Junru Lu, Yan Qian, H. Vicky Zhao, Di Yin, Xing Sun

Instrumentation and Methods for Astrophysics

pyEDITH: the coronagraphic exposure time calculator for the Habitable Worlds Observatory

To support the development of next-generation missions for the search and characterization of habitable planets, high-fidelity tools for astrophysical and instrumental noise simulations are needed. In this paper, we introduce pyEDITH, the Python-based coronagraphic exposure time calculator built for the next recommended NASA flagship mission, the Habitable Worlds Observatory (HWO), tasked with searching for signs of habitability and life in dozens of nearby exoplanet systems. pyEDITH is designed to simulate wavelength-dependent exposure times and signal-to-noise ratios (S/N) for synthetic HWO direct imaging observations, considering realistic engineering specifications and user-defined target information. Its modular architecture ensures flexibility as mission requirements evolve. pyEDITH enables a streamlined integration with modern astronomical workflows and was designed to be used by the scientific community at all skill levels for understanding the capabilities and limitations of different HWO architectures for exoplanet analyses. The code has been validated against existing exposure time calculators and released open-source on GitHub and Zenodo, as well as made accessible through a Graphical User Interface. The pyEDITH package includes API documentation, tutorial notebooks, and has been used in forthcoming scientific publications.
Read the full abstract

All authors Eleonora Alei, Miles H. Currie, Corey Spohn, Christopher C. Stark, Aki Roberge, Avi M. Mandell, Enrico Biancalani, Samantha Gilbert-Janizek, Jacob Lustig-Yaeger, Sarah Steiger

Computation and Language

ContextPilot: Teaching Agents for Proactive Context Management via Fine-grained RL

Long-horizon agentic tasks require large language models (LLMs) to iteratively retrieve, integrate, and maintain dispersed information across multi-turn interactions, but preserving all interaction histories leads to a continuously growing working context. Recent proactive context management methods allow models to edit their own working context with specialized tools, yet they still face three key limitations: (1) a limited toolset restricted to search, deletion, and summarization, with no support for global planning, long-term memory, and adaptive compression; (2) inefficient exploration that treats context management actions uniformly despite their heterogeneous impacts on final outcomes; and (3) coarse-grained credit assignment that assigns the final trajectory-level reward to all intermediate context editing actions during RL. To bridge these gaps, we introduce ContextPilot, a proactive context management framework for long-horizon agentic reasoning. Our approach systematically augments the toolset with planning, long-term memory, and soft context offloading tools. We further propose an RL method tailored for context management, which uses context and entropy variation to identify critical editing decisions for branch sampling and estimates action-level advantages from all branched trajectories that pass through the corresponding context editing action. Experiments on long-context QA and deep search tasks show that ContextPilot achieves stronger performance with a more compact working context, consistently outperforming existing baselines across various base models and benchmarks. Code is available at https://github.com/Tencent/ContextPilot.
Read the full abstract

All authors Zhuoshi Pan, Qizhi Pei, Junru Lu, Honglin Lin, H. Vicky Zhao, Di Yin, Xing Sun

Artificial Intelligence

COVER: Identifiable Evaluation of Coalition Routing

When a multi-agent system changes its team, it also changes the messages and final answer it produces, so an end-to-end accuracy gap does not by itself identify a routing effect. We introduce method, an evaluation contract that fixes a public information boundary, downstream stack G, and finite legal team family before outcomes are generated. Complete coverage identifies exact finite-benchmark oracle regret conditional on that stack. For any finite collection of frozen policies, executing the union of their distinct selected teams is the minimal assumption-free support for every pairwise policy contrast, though not for absolute oracle regret. Two controlled tables with source-ID-disjoint splits test the instrument. On MuSiQue-12, a pre-specified privileged positive control improves regret from 0.532 to 0.402; a later public-interface control reaches 0.424 versus 0.554 but is retrospective. On HotpotQA-4, a pre-specified public direct scorer improves regret from 0.313 to 0.110. In fixed-stack Llama execution, verified route regret improves by 0.190, while the raw-answer gain is 0.010 with an interval crossing zero. A five-family ToolSandbox variant-shift validation exhaustively evaluates 16 declared teams on 14 untouched task variants (224/224 valid rows): the declared-family oracle reaches 0.768 safe-evidence completion, while the prospectively frozen router gets 0.637 (regret 0.131), failing the predeclared 0.10 criterion. A later retrospective comparator reaches 0.655, matching all-workers with 4.57 versus 5.00 workers on average. Thus COVER exposes selection headroom without manufacturing a routing win. A crossed-stack diagnostic shows absolute scores depend on G but finds no detectable router-by-finalizer interaction. COVER is an auditable measurement methodology, not a claim of stack-invariant or universal agent-routing superiority.
Read the full abstract

Social and Information Networks

Structural Change and Random Graph Models in Global Oil Trade Networks

We studied structural change in global oil trade using a network approach. Using UN Comtrade data, we examined the temporal evolution of international trade networks, with an emphasis on crude oil. Weighted in-degree identified major changes in country rankings in 1991, 2011, 2017, and 2021, while PageRank detected pronounced changes around 1991 and 2024. The Louvain algorithm identified clear geographic communities within the overall trade network. In the oil trade network, modularity declined from the 1990s to the 2010s, with node2vec embeddings showing weaker clustering in 2011 than in 1991. We also compared the oil trade network with several random graph models using 3- and 4-node subgraph profiles and machine learning classification. The oil trade network was consistently classified as a Chung-Lu graph, while the Geometric model was not favored, suggesting that a model based on the degree distribution better matched its subgraph profiles than the other models considered.
Read the full abstract

Logic in Computer Science

On Left Adjoints Preserving Colimits in Homotopy Type Theory

We examine how the standard proof that left adjoints preserve colimits behaves in the setting of wild categories, a natural setting for synthetic homotopy theory inside homotopy type theory. We show that the proof may fail for adjunctions between wild categories and even produce a wild left adjoint that fails to preserve colimits. Our core contribution, however, is a sufficient condition on the left adjoint for the proof to go through. The condition, which we call 2-coherence, expresses that the naturality structure of the hom-isomorphism commutes with composition of morphisms. We present two useful examples of this condition in action. First, we use it, along with a new version of a known trick for homogeneous types, to show that the suspension functor, as well as a generalization thereof, preserves graph-indexed colimits. Second, we show that every modality, viewed as a functor on coslices of a type universe, is 2-coherent as a left adjoint to the forgetful functor from the subcategory of modal types, thereby proving this subcategory is cocomplete. We have formalized our main results in Agda.
Read the full abstract

Sound

Multirate State Space Models for End-to-End Processing of Pulse Density Modulated Speech Signals

Deep neural networks (DNNs) based on state-space models (SSMs) are increasingly applied to speech processing, but typically operate on pulse-code-modulated (PCM) audio. This constrains deployment on low-power, always-on edge devices, which commonly use single-bit pulse-density-modulated (PDM) micro-electromechanical (MEMS) microphones for their noise robustness, low cost, and variable sampling rates that enable low-power operation. In fact, converting PDM to PCM requires low-pass filtering and decimation, imposing costly overhead on resource-constrained hardware. While prior works have attempted to process PDM signals directly, they require long training times and generalize poorly across sampling rates. In this paper, we show that the SSM has two key properties that remediate these issues: its continuous-time parametrization allows it to produce a consistent representation of the input audio signal, regardless of the modulation strategy and sampling rate, and its long-term memory enables this representation to be aggressively downsampled without needing any anti-aliasing operations. We then propose a novel end-to-end PDM speech processing architecture that uses an SSM to encode the input audio signal into a modulation- and sampling-rate-invariant latent representation. We show that our proposed architecture achieves robust speech classification and enhancement gains at low-power sampling-rates (512 kHz) and similar performance to state-of-the-art algorithms operating on PCM data when tested on standard PDM sampling-rates of 2 MHz. Moreover, we show that the SSM's output can be downsampled by more than 65,000 times, thus significantly reducing the number of processing timesteps in downstream layers.
Read the full abstract

Algebraic Geometry

On Manin's conjecture for quartic del Pezzo fibrations

We generalize the homological sieve method, developed by Das, Lehmann, Tosteson, and the author, to study certain quartic del Pezzo fibrations, and we prove a version of Manin's conjecture over global function fields in these cases. Our proofs combine the $3$-dimensional positive-characteristic minimal model program, the geometry of the space of sections and the Abel–Jacobi mapping, and the homological sieve method. Our quartic del Pezzo surfaces have Picard rank $2$, and admit two birational morphisms to non-split quadric surfaces. In particular, they do not possess any conic fibrations.
Read the full abstract

Quantum Physics

Input-Output Analysis of Quantum Dot SUPER Excitation

The Swing-UP of the quantum EmitteR population (SUPER) two-color pulsed excitation scheme allows for robust and close to 100% excitation of a two-level quantum emitter using only red detuned light. We analyze the underlying counterintuitive dynamics using a full quantum input-output description of the in- and outgoing light pulses for the case of two coherent free-space input pulses at realistic photon numbers. At the microscopic level, the SUPER mechanism exhibits its nonlinear three-photon Raman-type character, leading to a net photon-number change of $-2$ in one mode and $+1$ in the other, as was first guessed from cavity-enhanced model descriptions at low photon numbers. We confirm that in free space, a sufficiently high pulse photon number, much larger than one, is required to achieve high-fidelity inversion. To treat the large coherent-state amplitudes (photon numbers) relevant for SUPER, we extend the quantum input-output formalism to include a cumulant expansion approach. With an interaction-picture formulation, the few exchanged photons that govern the nontrivial dynamics enable direct full-quantum calculations in truncated Hilbert spaces, including treatments in a displacement frame and for initial Fock-state pulses.
Read the full abstract

All authors Johannes Kerber, Christoph Hotter, Helmut Ritsch, Klaus Mølmer

Information Theory

Distributed Cross-Layer Optimization for Covert Multi-Hop, Multi-Modal Networks: Exponentially Fast Convergence and Robust Tracking

This paper develops the first distributed cross-layer algorithm for joint congestion control, routing, scheduling, and power control in covert multi-hop, multi-modal wireless networks, where adversarial wardens (Willies) monitor radio modalities via energy detection. The Detection Error Probability (DEP), the probability that a Willie fails to reliably detect ongoing transmissions, is generally non-concave in the transmit powers, making DEP-based covert network optimization challenging. We resolve this by constructing the tightest concave lower bound on the log-DEP, yielding a conservative convex problem that guarantees satisfaction of the original DEP constraints and unifies hard covertness constraints and covertness-utility maximization in a single problem. We develop a Parallel Proximal Alternating Direction Method of Multipliers (PP-ADMM) algorithm for the resulting cross-layer problem and prove global Q-linear convergence, i.e., exponentially fast convergence, to the set of optimal solutions under standard regularity conditions. Numerical results confirm linear convergence and demonstrate robust tracking performance under channel fading and Willie mobility.
Read the full abstract

All authors Sirin Chakraborty, Andrea Panebianco, Yuchen Tian, Kevin S Chan, Fikadu Dagefu, Yin Sun, Ness B. Shroff

Plasma Physics

Real-time virtual circuits for plasma shape control via neural network emulators: experimental demonstration on MAST Upgrade

Conventional plasma shape control in tokamaks relies on virtual circuits (VCs) that are computed offline from linearisations around a small, tailored number of reference equilibria, and deployed as expertly prepared schedules during the discharge. Here, we report on the first experimental deployment of real-time VCs. We replace pre-set look up tables with VCs updated in real time using surrogates of the plasma response. Both the existing control architecture and the interpretability of VC-based control are retained. Previous work showed that neural network emulators can produce accurate VCs, and validated their performance in closed-loop shape control simulations. Here, we report their first experimental validation on MAST Upgrade (MAST-U). Dedicated experiments spanning different scenarios, including prescribed shape perturbations, feedback-driven divertor-leg motion, and strongly evolving plasma configurations, show that real-time VCs can realise plasma shape control tasks within the MAST-U plasma control system. These results establish the experimental feasibility of real-time linearisations as a practical extension of conventional plasma shape control in tokamaks. The present implementation demonstrates a central step towards a simpler control workflow, in which manually constructed, phased VC schedules are replaced by VCs generated automatically online from a trained surrogate model, without scenario-specific retraining.
Read the full abstract

All authors Nicola C. Amorisco, Kamran Pentland, Adriano Agnello, George K. Holt, Alasdair Ross, Matthew J. Marshall, Edward Jones, Graham J. McArdle, Charles Vincent, Timothy Nunn, Martin Kochan, Pedro Cavestany, Aran Garrod, Stanislas Pamela, James Buchanan

Computation and Language

Stranger, Fan, or Peer? A Systematic Study on the Role of Interlocutor in Persona-Based Dialogue Generation

Persona-based dialogue systems are usually conditioned on speaker biography, but dialogues involve at least two participants, and who has access to whose biography can vary across training, inference, and evaluation. Prior work often neglected these aspects, obscuring mechanisms that only appear when biography visibility is toggled separately across training, inference, and evaluation, a three-stage factorisation that prior work has largely treated as a single factor. We study this factorisation on a dataset of dialogues paired with speaker's biographies, varying whether the target and interlocutor speakers see each other's biographies during training and inference, and using an LLM as a judge to perform author identification. We find that (i) training-time visibility, more than inference-time visibility, determines whether models express persona traits through dialogue or fall back on copying biographical text (a known problem/phenomenon in persona-based generation); (ii) models trained with interlocutor-biography visibility copy less target-biographical text than models trained without it, while changing visibility only at inference time has a less consistent effect; and (iii) under asymmetric disclosure, where only the interlocutor sees the target biography, target content leaks into interlocutor turns more often, and dialogues containing such traces are easier for the judge to identify, especially when interlocutor turns are visible. These results suggest that biography leakage into generated turns is an artefact of how interlocutor visibility is configured across training and inference, and separating the three stages is necessary.
Read the full abstract

Applied Physics

Synergy between laser linewidth and frequency chirp in mesospheric magnetometry based on the sodium laser guide star

Mesospheric sodium magnetometry with a laser guide star measures the geomagnetic field near 90~km. Its sensitivity hinges on laser linewidth and chirp, yet prior work optimized these two parameters only separately. We use velocity-resolved density-matrix simulations of Larmor-synchronous pulsed Na D$_2$ pumping to scan both parameters jointly. Linewidth and chirp exhibit a synergy: when chirping carries the recoil-mitigation role, the return flux stays within 1% of its peak across linewidths of 2–10~MHz. The flux-optimal chirp is $0.19~\mathrm{MHz/\upmu s}$, one fifth of the continuous-wave rate; this synergy guides high-sensitivity mesospheric magnetometer design.
Read the full abstract

All authors Yucheng Yang, Chunyang Lei, Kai Guo, Shuai Wang

Algebraic Geometry

Betti bounds for spaces of curves on varieties and Manin's conjecture for quartic del Pezzo surfaces

We prove uniform exponential bounds for the compactly supported Betti numbers of spaces of morphisms from curves of fixed genus to projective varieties. For targets in a fixed projective space cut out by a prescribed number of equations of fixed degrees, the bound is exponential in the degree of the morphism and is independent of the ground field, the source curve, and the target. The proof constructs bounded-degree affine presentations involving only linearly many variables and equations, and then applies Katz's estimate. As an application, we establish a higher genus function field version of Manin's conjecture for split quartic del Pezzo surfaces, generalizing a recent theorem of Das–Lehmann–Tanimoto–Tosteson. The passage from $\mathbb{P}^1$ to arbitrary source curves requires uniform bounds for compactly supported Betti numbers over all curves of fixed genus, together with new geometric results on higher genus morphism spaces, including irreducibility and dimension estimates for the relevant incidence strata. These results, combined with the necessary control of the configuration space contributions, allow the homological sieve and virtual height zeta function argument to yield, over sufficiently large finite fields and within a slightly shrunken nef cone, the predicted asymptotic with the expected leading constant.
Read the full abstract

High Energy Physics (Phenomenology)

Impact of NLO EW and QCD corrections to dimension-6 SMEFT coefficients constraints in Higgs decays

This paper presents the impact of the inclusion of next-to-leading-order (NLO) QCD and electroweak (EW) corrections for dimension-6 Standard Model Effective Field Theory (SMEFT) predictions for the Higgs decays in Higgs couplings measurements at the Large Hadron Collider. Such higher-order corrections will be increasingly important for the precise Higgs and electroweak measurements expected at the High-Luminosity Large Hadron Collider (HL-LHC). The results are compared with a SMEFT parameterisation employing inputs from a previous ATLAS study, in which the dimension-6 SMEFT contributions are evaluated for each production and decay process at the lowest non-vanishing perturbative QCD order of the corresponding SM process. The effects of the SMEFT NLO QCD/EW contributions to the Higgs decay channels are largest for 2-fermion operators involving the third generation. The sensitivity to the Higgs tri-linear coupling is also explored.
Read the full abstract

All authors Luigi Bellafronte, Ana Rosario Cueto Gómez, Sally Dawson, Clara Del Pio, Matthew Forslund, Pier Paolo Giardino, Andrea Visibile

Quantum Physics

Fabrication-free assessment of microwave losses in germanium-based dielectrics and superconductors

We present a flip-chip-based sensing scheme to measure effective microwave losses associated with target materials for quantum technologies, without requiring any device fabrication on the material under test. Using this approach, we quantify the microwave losses of a strain-engineered Ge/SiGe quantum well heterostructure and investigate losses arising from its Ge substrate and intermediate layers. The quality factors of the fabricated microwave resonators agree with the losses of dielectric materials independently extracted from flip-chip sensing measurements. We further study the superconductor platinum silicon germanide (PtSiGe) prepared by thermal reaction with a deposited Pt film, finding high microwave losses that limit the suitability of the films studied here as the sole superconductor for high-quality resonator applications. By coating Pt with Nb prior to the reaction, we observe a substantial reduction in microwave loss and a nearly three-fold enhancement of the transport critical temperature. The temperature dependence of the microwave loss is consistent with gap inhomogeneity in both superconducting films. These results identify constraints on material choices, provide design guidance for microwave circuits on planar Ge heterostructures, and demonstrate a fast-turnaround testing method for new materials for superconducting quantum circuits.
Read the full abstract

All authors Haoran Lu, Kushagra Aggarwal, Xiangqin Wang, Pauline Drexler, Daniel Tong, Maciej W. Olszewski, Anand Ithepalli, Lingda Kong, Debdeep Jena, Peter L. McMahon, David A. Muller, Dominique Bougeard, Valla Fatemi

Data Structures and Algorithms

Tight Bounds for Memory Allocation With and Without Request Fragmentation

The classical memory-allocation problem captures the task of placing objects of different sizes in memory, while minimizing the so-called memory high-water mark. It has been known since the early 1970s that the optimal competitive ratio for any deterministic online allocator is $\Theta(\log M)$, where $M$ is the volume high-water mark of the underlying request sequence. This paper begins with a simple observation: many real-world allocators seem to bypass the 1971 lower bound by adopting a slightly different model for memory allocation. These allocators use what we call $k$-aggregate request fragmentation, meaning that the memory allocator is permitted to break requests into multiple fragments, so long as the all-time maximum number of simultaneous fragments is at most $k$ times the all-time maximum number of simultaneous requests. We consider the following basic question: Does request fragmentation fundamentally change the problem of memory allocation, and if so, how? Our results come with several surprises. Among these, we find that even using $k = 1 + o(1)$ request fragmentation, the optimal competitive ratio—which was $\Theta(\log M)$ in the classical setting—collapses to $\Theta(\log \log M)$. This result is shown to be tight with matching upper and lower bounds, applying to both deterministic and randomized algorithms.
Read the full abstract

All authors Michael A. Bender, Alex Conway, Martín Farach-Colton, Hanna Komlós, William Kuszmaul, Nicole Wein

Computer Vision

Anatomy-Aware Promptable Segmentation with Online Interactive Training for AUTOPET V

We present an anatomy-aware, promptable model for whole-body lesion segmentation in FDG and PSMA PET/CT, developed for the AUTOPET V challenge. The proposed method is built as family of nnU-Net-based models and trained in two stages: i) a pre-training stage that produces a strong initial segmentation, and ii) an online interactive stage that learns to exploit scribble prompts, refining the prediction over successive interactions. Anatomical context is incorporated through organ supervision using a single shared head that predicts lesions and organs from the same features, which reduces false positives arising from physiological uptake. Also as the tracer (i.e., FDG/PSMA) is not provided at inference, we add a tracer classifier based on image processing and a random forest over coronal MIP features, routing each study to a combined FDG+PSMA model or to a PSMA-specific model. Across four-fold cross-validation the organ-supervised model achieves the best and most stable performance, the interactive stage improves the Dice score monotonically with each prompt, and PSMA-specific training yields the strongest tracer-wise results.
Read the full abstract

Computer Vision

LayerRecall: A State-Conditioned Memory Router for Long-Horizon Consistency in Video Generation

Autoregressive video diffusion enables scalable long-video generation by producing chunks from a bounded recent context. While recency-based caching preserves local continuity, it evicts historical cues needed when subjects, objects, scenes, or attributes reappear. Existing memory mechanisms expose models to nonlocal history, but access alone does not ensure effective use. Our analysis reveals that video DiT layers exhibit distinct preferences for current, recent, and distant context, suggesting that long-range memory requires deciding both what to retrieve and where to use it. We introduce LayerRecall, a current-conditioned, layer-selective memory router that retrieves relevant historical K/V states and injects them only into backbone-specific memory-sensitive layers while preserving local attention elsewhere. To reduce reliance on scarce high-quality long-horizon videos and explicit memory-allocation labels, we further propose Cross-Horizon Prediction Matching (CHPM), which uses a privileged long-context reference to supervise the bounded-memory router in prediction space. Across 100 multi-shot evaluation prompts, LayerRecall achieves the best overall results on MemoBench and MovieBench while matching its backbone on VBench-Long, demonstrating stronger long-range recovery without sacrificing local continuity. Qualitative analyses further reveal memory-guided self-correction, whereby initially mismatched local attributes return to their historical appearance without resetting ongoing motion or scene structure. Additional analyses show cross-backbone portability and negligible inference overhead.
Read the full abstract

All authors Yixuan Ding, Jiahao Kong, Wei Huang, Ruijie Quan, Yi Yang

Analysis of PDEs

Spectral gap for the three-dimensional damped cubic wave equation with degenerate noise

We establish a weighted Wasserstein spectral gap for the three-dimensional damped cubic wave equation with genuinely finite rank Brownian forcing. Under a saturation condition, the gap holds with respect to the negative phase topology $\mathcal E_s=H^{-s}\times H^{-1-s}$ for every $0<s<1/2$, from which we deduce unique ergodicity and exponential mixing in the energy topology. Sharp geometric characterizations of the saturation condition are also obtained. The method we develop is a stable–compact asymptotic coupling mechanism for hypoelliptic dissipative SPDEs beyond the parabolic setting. Instead of relying on positive time smoothing or asymptotic gradient estimates, it reduces the infinite-dimensional obstruction to contraction to a compact defect in a weaker coupling topology. Dense Malliavin range then permits this defect to be compensated by a finite dimensional Cameron–Martin shift, producing a finite distance contraction on bounded Lyapunov cores. Together with a separate high energy contraction from dissipation, this yields a global weighted Wasserstein spectral gap.
Read the full abstract

Computation and Language

Acquire, Repair, Preserve: A Diagnosis-Guided Post-Training Recipe for Small-Model Dialogue Game Agents

Interactive dialogue games test a capability that static benchmarks largely leave implicit: a model must carry state across turns, interpret feedback, and choose valid actions under changing constraints. We study this setting in the LM Playschool Challenge with a 2B open-weight model, and find that many failures are not only broad knowledge failures but also local decision failures: repeated guesses, malformed actions, and violations of feedback that the model has just seen. These diagnostics motivate a training recipe organized around three steps: acquire broad game participation through supervised fine-tuning, repair mechanically verifiable failures within one targeted dialogue-game family using turn-local preference pairs, and preserve general capabilities beyond these dialogue games. In the official final evaluation, our submission improves public clemscore from 10.67 to 38.92 and closed in-domain score from 13.41 to 41.17, while approximately preserving aggregate static performance (44.14 vs. 44.24 for the baseline). Out-of-domain clemscore remains low at 7.88, with the largest gains concentrated in unseen variants of the targeted family. Our results suggest that broad SFT brings most of the model's capability improvement; turn-local supervision can be effective when failure detection is precise, with observed transfer concentrated primarily within-family.
Read the full abstract

Other Condensed Matter

Magnonic Combinatorial Memory based on a network of coupled active ring circuits

Magnonic Combinatorial Memory (MCM) is a type of memory where the bits of information are encoded in the signal propagation paths in the network. In this work, we consider MCM based on the network of a coupled active ring circuit (ARC). Each circuit includes a broadband amplifier, a magnonic delay line, an adjustable frequency filter, an adjustable phase shifter, and a power detector. The coupling between the circuits is via spin waves propagating in the common delay line - ferrite film. There may or may not be auto-oscillations in the active ring circuits, depending on the combination of circuit parameters and circuit coupling. The address of MCM is defined as the combination of the states of the phase shifters and frequency filters, while the MCM state is defined as the presence/absence of the auto-oscillations. The coupling between the circuits is achieved by placing micromagnets on top of the ferrite film. The number of bits that can be encoded in the network increases quadratically with the number of coupled circuits. This scaling provides a fundamental advantage over conventional memory. We present experimental data obtained for three magnonic ARCs connected via a single-crystal yttrium iron garnet Y3Fe2(FeO4)3 (YIG) film. The data illustrate an example of encoding a 27-bit binary response pattern, corresponding to the 27 experimentally accessible phase combinations. The results demonstrate a robust operation of MCM with an On/Off ratio exceeding 30 dB at room temperature. The advantages and shortcomings of the proposed approach are discussed.
Read the full abstract

All authors Mykhaylo Balinskiy, Paulo Julio, Jeffrey Vargas, Diana Bisono Balaguer, Jacob Greenstein, Alexander Khitun

High Energy Physics (Phenomenology)

Effect of Thermal Broadening on Light Hadron Production in Heavy-Ion Collisions

The measured hadron yields in heavy-ion collisions are well described by thermal particle production at temperatures close to the QCD pseudo-critical transition. However, if particles are produced in thermal equilibrium, then the in-medium effects, e.g., broadening of their spectral functions, must be taken into account. In this work, we study for the first time the effect of the thermal broadening of the light vector-meson spectral functions on light hadron production in heavy-ion collisions. We generalize the efficient resonance decay framework FastReso to resonances with finite-width spectral functions. We then compute the pion, kaon and proton spectra from the decays of hadron resonances for in-medium and vacuum spectral functions. We perform a simultaneous blast-wave fit with decay feed-down to measured particle spectra in central Pb-Pb and Xe-Xe collisions at the LHC. We find an increase in pion spectra at low momenta, from including thermally broadened $\rho$ and $\omega$ mesons. Additional pion yield from finite-resonance widths reduces the commonly observed discrepancy between models and data, but does not completely resolve the soft pion puzzle in heavy-ion collisions.
Read the full abstract

Computer Vision

ARC-CT: Anatomy-Routed Contrastive Vision-Language Learning for 3D Chest CT

Contrastive vision-language learning uses paired chest CT volumes and radiology reports to learn abnormality classifiers without manually annotated labels. However, two characteristics of chest CT challenge conventional global contrastive learning. First, many critical abnormalities are small or anatomically localized, and pooling an en- tire volume into a single embedding may dilute their visual evidence. Second, the standard contrastive objective treats every other scan in a batch as a negative. Because many chest CTs share abnormalities, this objective incorrectly pushes co-positive pairs apart. We propose Anatomy-Routed Contrastive Learning for 3D Chest CT (ARC-CT), a region-aware framework that addresses these limitations using only la- bels extracted from reports by an LLM, with no manual annotations or bounding boxes. ARC-CT combines three components: (1) an Anato- myQFormer localizing evidence via queries constrained by automatically generated organ masks; (2) a label-Jaccard soft InfoNCE objective in- tegrating the standard one-hot target with the label-set overlap of each pair, which reduces false-negative penalties between studies that share clinical findings; and (3) an organ-level alignment loss connecting mask- pooled visual features to organ-specific report text extracted offline with a large language model. ARC-CT achieves a 0.86 mask-free macro AUC across 18 abnormalities using a compact 3D ResNet-18 backbone. Over- all, ARC-CT outperforms both comparable efficient baselines and sev- eral larger transformer models. Our code and weights are available at https://github.com/arc-ct/arc-ct.
Read the full abstract

All authors Huseyin Umut Isik, Mehmet Alp Ozaydin, Sila Kurugol, Şeyda Ertekin

Optics

Correlation Swapping: a correlator for independent thermal light sources

Optical intensity correlation is a fundamental property of light and an essential resource for numerous optical applications. In this work, we introduce the concept of classical correlation swapping, a classical analogue of quantum entanglement swapping, to generate all-purpose spatial correlations between two independent thermal light sources. Using a spatially unresolved Mach-Zehnder interferometer and a medium variable, we theoretically and experimentally demonstrate the feasibility of the correlator with two distinct schemes for independent pseudo-thermal light beams. Notably, the resulting photon correlations can be readily tailored to exhibit either bunching (peak) or anti-correlated (dip) characteristics, with a significantly reduced number of post-selection measurements. Leveraging this classical correlator, we further demonstrate the first classical ghost imaging experiment using uncorrelated or unknown light. Numerical simulations also confirm that the correlator can not only operate well in other spatial degrees of freedom (e.g., orbital angular momentum) but also be used to establish specific spatial correlations between pseudo-thermal light sources possessing distinct correlation properties. This work opens a new avenue for harnessing uncorrelated classical light in correlation-based optical applications.
Read the full abstract

All authors Wanting Hou, Jilun Zhao, Zhiyuan Ye, Hong-Chao Liu, Hai-Bo Wang, Jun Xiong, Kaige Wang

Computer Vision

Prompt-Guided Interactive Segmentation of Interstitial Lung Disease in Thoracic CT

Accurate segmentation of interstitial lung disease (ILD) patterns is essential for quantitative disease assessment and longitudinal monitoring. However, existing approaches remain limited by relying on dense annotations and producing static predictions that cannot be refined, motivating interactive approaches. While promptable models show promise in interactive segmentation, their adaptation to ILDs remains largely unexplored. To address this gap, we investigate prompt-guided foundation models for ILD refinement and present, to the best of our knowledge, the first adaptation of MedSAM2 for interactive 3D ILD segmentation on thoracic CT. We investigate three fine-tuning strategies and multiple clinically motivated prompts: bounding-boxes (BBox), point, lasso, and scribble. On a dataset spanning seven ILD patterns and healthy lung tissue, full model fine-tuning performed best, improving the average Dice score by 4.7 percentage points over MedSAM2.While BBox prompts achieve the strongest performance, non-native MedSAM2 interactions such as lasso and scribble prompts also prove effective. Finally, we present and evaluate a proof-of-concept end-to-end workflow in which MedSAM2 is initialized from an automatic segmentation prior and subsequently refined using radiologist prompts. Model weights and plug-ins made available at: https://github.com/AIHNlab/ILD-SemiSegTool.
Read the full abstract

All authors Vasilis Dedousis, Lubnaa Abdur Rahman, Lorenzo Brigat{\omicron}, Ethan Dack, Andreas Christe, Christoph Frank, Manuela Funke-Chambour, Justus Roos, Adrian Huber, Lukas Ebner, Stavroula Mougiakakou

Combinatorics

An Exposition of the $\widetilde{O}(\log^{1/4} n)$ Bound for the Komlós Problem

A conjecture of Komlós states that the combinatorial discrepancy of any matrix $A\in\mathbb R^{m\times n}$ whose columns have Euclidean norm at most one is bounded by a universal constant. We prove that the combinatorial discrepancy of every such matrix is at most $O((\log n)^{1/4}(\log\log n)^{7/4})$. This is the first asymptotic improvement over the $O(\sqrt{\log n})$ bound established by Banaszczyk [Banaszczyk, Random Struct.\ Algorithms, 1998], and it refutes a conjecture of Hajela [Hajela, European J.\ Combin., 1988] that a lower bound of order $\Omega(\sqrt{\log n})$ should hold.
Read the full abstract

Quantum Physics

Constant-time equilibration of observables under rapid Lindbladian dynamics

Markovian open-system dynamics have widespread applications throughout quantum information science, including algorithmic state preparation. Their convergence is commonly quantified using the worst case global trace distance between the evolving and stationary states. However, this criterion can be unnecessarily stringent when only physically relevant observables are of interest. Here we introduce and study observable-specific mixing times. We prove that, for quasi-local, rapidly mixing Lindbladians, sums of geometrically local observables equilibrate in a time independent of system size, in contrast to the logarithmic dependence of global state mixing. This separation reduces the runtime of dissipative quantum algorithms, including quantum Gibbs samplers, for estimating quantities such as the Gibbs state energy and local order parameters, yielding an overall scaling that is linear in system size. Complementing this quantum result, we develop a quantum-inspired classical algorithm for estimating the same quantities. Its runtime is likewise linear in system size, but scaling exponentially in $\mathcal{O}\big(\log(1/\epsilon)^D\big)$, where $D$ denotes the spatial dimension of the lattice. We further analyse non-interacting Lindbladians over qudits, fermions, and bosons, demonstrating that locality of observables is not always necessary for a qualitatively faster mixing. Small-scale simulations of quantum Gibbs samplers reveal no large hidden constants in our asymptotic analysis and show that the theoretical predictions closely capture the finite-size dynamics.
Read the full abstract

All authors Štěpán Šmíd, Richard Meister, Mario Berta, Roberto Bondesan

Analysis of PDEs

Propagation rates in integro-differential equations of ignition type

This paper provides sharp rates of invasion in nonlinear integro-differential equations of ignition type. The framework covers both integrable convolution kernels and singular fractional-type kernels. While the emergence of a threshold on the decay $1+2s$ of the jump kernel is expected from earlier works, far fewer quantitative results were available in such a general framework. In particular, in addition to the spreading rates, our study provides information on the shape of the invasion profiles. Importantly, we obtain the first sharp spreading estimate in the critical case $s=\frac12$, which separates linear-in-time and accelerated regimes.
Read the full abstract