Machine Learning
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Tuesday 1 September
Black-Box Optimization (BBO) has broad applications, while traditional algorithms such as evolutionary algorithms and Bayesian optimization face efficiency challenges as real-world BBO problems grow increasingly complex. Data-driven optimization has been the most popular paradigm to improve the efficiency of BBO, by learning from data. Offline data-driven optimization seeks high-quality solutions using only a fixed set of previous evaluations, attracting substantial attention because it requires no additional online evaluations. Many offline optimization methods have been proposed, but a fundamental question remains unanswered: what learnability is sufficient for offline optimization? Prior theoretical studies show that Probably Approximately Correct (PAC) learnability is insufficient, as the optimal region may remain poorly learned even when most regions are well learned. In this paper, we propose algorithm-dependent learnability, which requires accuracy only on the optimizer's trajectory. We prove that its value-query form is sufficient for representative discrete settings, including greedy and local search for submodular maximization, while its first-order analogue is sufficient for projected gradient descent on convex minimization. Motivated by this notion, we formalize a trajectory-learning framework comprising trajectory construction, trajectory modeling, and candidate generation, and analyze existing trajectory-based methods under it. We further propose Uncertainty-aware Gradient-guided Trajectory Learning (UGTL), which constructs locally coherent improvement trajectories reflecting plausible search paths, models them with conditional diffusion, and selects a diverse candidate set. Our experiments show that UGTL achieves the best average rank, 3.1/25, among 25 methods on Design-Bench tasks, and confirm that our trajectory construction plays a significant role in the improvement.
Materials Science
Accelerating dynamic simulations of photoexcited materials and their evolution by electron-informed machine learning
Nonadiabatic coupled electron-nuclear dynamics upon electronic excitation underpin the microscopic mechanism and rational modulation of diverse photoinduced functional phenomena in materials, yet their direct first-principles simulations remain computationally demanding. Here we develop a framework for nonadiabatic excited-state machine-learning molecular dynamics (EMLMD) simulations, where the nonequilibrium electronic information upon photoexcitation such as electron temperature is rigorously calibrated from high-precision real-time time-dependent density functional theory (rt-TDDFT) benchmark simulations, enabling accurate reconstruction of excited-state potential energy surfaces (PES). This framework natively incorporates the excited-state electron-phonon couplings and intrinsically captures photoinduced phonon anharmonicity, both of which are missing in standard machine learning molecular dynamics, thus delivering first-principles-level accuracy for excited-state atomic evolutions. Large-scale EMLMD simulations resolve time- and momentum-resolved phonon dynamics in photoexcited materials, directly uncovering the competition between photogenerated coherent phonons and thermal phonons during photoinduced phase transition of bismuth. It also simultaneously resolves elusive atomic-scale microscopic dynamics and global structural rearrangement for selenium photoamorphization. Balancing high accuracy and efficiency, EMLMD offers a versatile paradigm to tackle key challenges in the study of complex excited-state molecular dynamics.
Read the full abstract
Computation and Language
GlossoGen: Emergent Language in Complex Multi-Agent LLM Interactions
The growing rate at which LLM agents interact with one another raises key questions about language evolution in multi-LLM-agent settings, with implications for safety and monitorability as well as for linguistic accounts of LLMs. To address these questions, we introduce GlossoGen, a novel platform for studying multi-agent language evolution in complex scenarios. Within GlossoGen, we build the SaveVeyru scenario, which requires agents with partial information to communicate under pressure. We find that language evolution does occur between LLM agents, that the resulting languages are compositional and morphologically productive, and that they deviate from the LLMs' English prior in ways that render them incomprehensible to humans. Moreover, we identify several qualities essential to this evolution: pressure towards efficiency; the strength of the models backing the agents; and access to a "postmortem" stage in which agents can agree on linguistic conventions. Importantly, we observe that different conditions govern the transmission of language to new agents. Specifically, we find that agents learn new languages from usage alone, take an active role in this learning, and that while stronger models are required for novel language emergence, weaker models can learn an existing language once it has emerged. Taken together, our results indicate that current LLMs have the potential for cumulative cultural evolution – previously attested only in humans – with mixed populations of agents developing capacities that go beyond their lowest common denominator.
Read the full abstract
Distributed Computing
Faster Convergence of Multidimensional Approximate Agreement via Smallest Enclosing Balls
This work considers the multidimensional approximate agreement problem. In this problem, $n$ parties in a distributed system, up to $t$ of which may be corrupted by a Byzantine adversary, need to output vectors that are close to each other and that lie inside the convex hull of all non-corrupted input vectors. We assume that nodes communicate in a fully-connected authenticated network and analyze synchronous and asynchronous communication models. The focus of this work is on the contraction factor of approximate agreement protocols. The first multidimensional approximate agreement protocols had a contraction rate of $1-1/n$(VG, PODC'13) and $\sqrt[d]{1/2}$(MH, STOC'13). While a rate below $1$ is sufficient for convergence, it is not sufficient for practical applications. To date, the best known convergence rate of approximate agreement algorithms is $\sqrt{7/8}\approx0.935$ (FN, DISC'18), which is achieved through the MidExtremes protocol. This stands in contrast to the lower bound on the convergence rate in the $1$-dimensional setting, which is $1/2$. In this work, we propose BallMidpoint - a novel approximate agreement protocol with a contraction rate of $1/\sqrt{2}\approx 0.707$ in the synchronous and the asynchronous communication models. This algorithm satisfies the optimal resilience under convex validity. The presented contraction rate is achieved by choosing the midpoint of the smallest enclosing ball of the so-called local safe areas, and it is tight for the presented algorithms. Similar to (FN, DISC'18), our algorithm is coordinate-free, and the point inside the safe area can be computed efficiently. This work presents the first multidimensional approximate agreement protocol where the convergence rate is closer to the best known lower bound rather than the upper bound of $1$.
Read the full abstract
Theoretical Economics
Bi-Compositional Division Rules
We characterise the division rules for claims problems that satisfy equal treatment of equals, bilateral consistency, composition down, and composition up. The rules are precisely the members of a one-parameter log-exponential family $\{r^\theta\}_{\theta\in[-\infty,+\infty]}$, with constrained equal awards (CEA) and constrained equal losses (CEL) as its endpoints. For finite $\theta$, $r^\theta$ is the equal-sacrifice rule in awards for $u_\theta=\log\varphi_\theta$, where \[ \varphi_\theta(x):=\frac{e^{\theta x}-1}{\theta}\quad(\theta\ne0), \qquad \varphi_0(x):=x, \] and simultaneously the equal-sacrifice rule in losses for the dual utility $u_{-\theta}$. The proportional rule is the midpoint, $\theta=0$. Continuity of the rules are not assumed but a consequence of the other axioms.
Read the full abstract
Analysis of PDEs
Stabilization by a background magnetic field: global well-posedness of the full compressible viscous non-resistive MHD system without heat-conductivity
We consider the three-dimensional full compressible magnetohydrodynamic(MHD) system on the periodic torus $\mathbb T^3$ in the regime where the only dissipative mechanism acting on the system is the viscosity of the fluid: the magnetic field is non-resistive and the flow is non-heat-conducting. We prove that this system admits a unique global smooth solution, together with explicit algebraic decay rates, provided that the perturbation $(\mathbf u_0,\,P_0-\bar P,\,\mathbf H_0-\mathbf n)$ of the equilibrium state $(\mathbf 0,\bar P,\mathbf n)$ is sufficiently small in a high-order Sobolev space and the background magnetic field $\mathbf n\in\mathbb R^3$ satisfies a Diophantine condition. No smallness whatsoever is imposed on the initial density: it is only required to be bounded away from vacuum and from infinity, and may exhibit arbitrarily large variations. The proof uncovers a hidden dissipation mechanism. Although neither the density, nor the pressure, nor the magnetic field is endowed with any diffusion or damping of its own, the coupling of these quantities with the velocity through the background field $\mathbf n$, combined with a Poincaré-type inequality of Diophantine origin, generates effective dissipation for both the pressure and the magnetic field perturbations. The large variations of the density are handled by a two-tier energy argument, in which weighted time-decay estimates for the intermediate-order energy compensate exactly for the linear-in-time growth of the highest-order norm of the density.
Read the full abstract
Cryptography and Security
Defense-as-Skill: Evolving Runtime Guard Skill for Skill-Augmented Agents
Skill-augmented agents load reusable skills as persistent runtime context, improving task performance but also giving malicious skills a durable channel for steering future actions. Such skills may leak secrets, corrupt code, bypass approvals, or stage data for exfiltration only after a concrete user task and workspace state make the unsafe action appear useful. This makes pre-install vetting insufficient and calls for runtime, task-conditioned protection. We propose Defense-as-Skill, a defense paradigm that implements the runtime guard itself as an installable, inspectable, and editable skill. Our guard, SkillSonar, runs alongside untrusted task skills and checks sensitive actions against the user's task boundary, routing each action to an allow, replan, or confirmation decision without modifying the underlying agent runtime. To study this setting, we construct SCOPE-R, a task-conditioned dataset covering 6 risk families and 21 sub-categories, with 206 attack-confirmed malicious instances and 43 benign tasks. We then improve SkillSonar on the SCOPE-R training subset using runtime guard-skill evolution, a Monte-Carlo Tree Search procedure that evolves the on-disk guard skill from feedback on the rollouts. Across Claude Code and OpenClaw, the evolved guard substantially reduces attack success while maintaining a favorable safety-utility trade-off; on repeated GLM-5 runs, ID ASR drops from 0.482 to 0.104 and OOD ASR from 0.606 to 0.115, with gains transferring across victim models, held-out risk families, and external benchmarks. These results show that runtime guards can be skill-native, transparent, and evolvable policy artifacts for skill-augmented agents.
Read the full abstract
Combinatorics
Induced Embeddings of Graphs into Abelian Cayley Graphs
For a finite graph $G$ on $n$ vertices, let $\eta(G)$ denote the least order of a finite abelian group $\Gamma$ for which $G$ is an induced subgraph of some Cayley graph of $\Gamma$. Babai and Sós (1985) settled the worst-case order of magnitude: it is $\Theta(n^2)$. We treat $\eta$ instead as an invariant of the individual graph, minimised over all finite abelian groups rather than over the cyclic groups alone, which is the restriction implicit in the literature on representation numbers modulo $n$. We prove a local order floor: $\eta(G)$ is at least the maximum of $n$ and twice the largest independence number of a neighbourhood of $G$. This localises at an arbitrary vertex the correspondence of Babai and Sós between induced stars and sum-free sets; a corollary of the classification of maximum sum-free sets in abelian groups does not lower this floor, but restricts which host orders are admissible and so prunes the search. We determine $\eta$ exactly for paths, where it equals $n+1$, and for complete bipartite graphs $K_{a,b}$, where it equals $2\max(a,b)$ and meets the floor. A Cartesian product bound gives $\eta(P_m \,\square\, P_m) = (1+o(1))n$. We report certified exact values of $\eta$ for $22$ graphs, computed over all abelian groups. Seventeen of the $22$ optimal hosts are cyclic, so on most of these graphs the cyclic restriction costs nothing; where it bites, however, it is expensive. A search restricted to cyclic groups returns $36$ for the Petersen graph against the true value $16$, and $59$ for the Frucht graph against $27$. The cost of the restriction is concentrated rather than diffuse, and we identify the graphs on which it is paid. We also determine $\eta$ exactly for the double stars $D_{q,q}$ with $2 \le q \le 6$, obtaining $5q$ in each case. Since $\eta(D_{6,6}) = 30$ exceeds $2n = 28$, no constant below $15/7$ can bound $\eta(T)/n$ over all trees.
Read the full abstract
Number Theory
On the Ramification of Quadratic Fields where there is a torsion growth
Let $E/\mathbb{Q}$ be a rational elliptic curve whose torsion group grows over a quadratic field $K$. In a previous paper, a relation between the primes of the conductor of $E$ and the primes that divide the discriminant of $K$ was shown. In the present paper, we go further in this study and we compute the ramification of the field $K$, when a torsion point of prime order $\ell$ is added, in terms of invariants of the curve. Concretely, the conductor, the prime $\ell$ and the Kodaira symbol at the primes of bad reduction.
Read the full abstract
Number Theory
A General Construction of Codes from Drinfeld Modules
We construct additive rank-metric and sum-rank-metric codes from Drinfeld modules by restricting bounded-degree morphisms to prime-to-characteristic torsion. For supersingular Drinfeld modules of rank $r$ in characteristic $\mathfrak{p}$ of degree $d$, the stabilization formula for morphism spaces yields rank-metric codes of $\mathbb{F}_q$-dimension $mrt-c$ and minimum distance $r-t+1$, where $c=r(r-1)(d-1)/2$. Simultaneous restriction to $\ell$ distinct degree-$m$ torsion modules gives additive sum-rank codes of the same dimension and minimum distance at least $\ell r-t+1$. Their normalized Singleton defects tend to zero, while in characteristic $(T)$ the module $\phi_T=\tau^r$ makes the defect vanish and produces an explicit MSRD family. We identify this family with a skew Chinese remainder theorem code supported on central skew polynomials and prove that its poly-skew weight is exactly $m$ times its sum-rank weight. This gives a specialized Singleton-type bound and a polynomial-time unique decoder up to the full sum-rank unique-decoding radius. We also derive a Welch-Berlekamp-type filter equation for the general supersingular sum-rank construction; it becomes an effective decoder whenever bases of the relevant morphism spaces and the restriction maps are computable.
Read the full abstract
Combinatorics
On cancellative pairs of families of subsets
A pair $(\mathcal{A}, \mathcal{B})$ of families of subsets of $[n]$ is cancellative if whenever $A, A' \in \mathcal{A}, B \in \mathcal{B}$ satisfy $A \cup B=A' \cup B$, then $A=A'$, and whenever $A \in \mathcal{A}, B, B' \in \mathcal{B}$ satisfy $A \cup B=A \cup B'$, then $B=B'$. We show that for every cancellative pair $(\mathcal{A}, \mathcal{B})$, the inequality $|\mathcal{A}||\mathcal{B}| \le 2.25^n$ holds, matching Tolhuizen's $(2.25-o(1))^n$ lower bound construction.
Read the full abstract
Fluid Dynamics
Jet creation and wave energy dissipation by a submerged elastic plate
Experiments on a flexible plate horizontally submerged in a wave tank are reported in this paper. The plate is held at a fixed depth at its upstream end, while its downstream end is free to move. The wave field is measured using a synthetic Schlieren technique, giving access to an estimate of the fractions of the incoming wave energy that are transmitted and reflected after encountering the region where the plate is submerged. Part of the incoming wave energy is shown to be dissipated by the plate. Flow field measurements in a vertical plane parallel to the wave propagation are performed to examine the wave-plate interaction. A jet-like flow is produced at the free end of the plate, with strong vortex activity similar to that produced by a flapping fin. The kinetic energy present in this jet-like flow is computed from the velocity field measurements, and the external dissipation produced by jet creation is estimated. Results highlight the efficiency of flexible plates in reflecting and attenuating waves. This study also showcases a change in jet direction, in particular the observation of a jet going upward toward the free surface for certain frequencies. Such a configuration, could be of great interest for wave energy conversion applications, while preventing harmful phenomena such as seabed erosion.
Read the full abstract
Artificial Intelligence
Harness-of-Harness: Multi-Day Autonomous Software Development with Continual Improvement
This paper studies autonomous software development, in which LLM-based coding agents transform high-level requirements into complete, functional, and usable software systems without human intervention. We introduce Harness-of-Harness (HoH), a framework that enables coding agents to continually improve software during autonomous development. HoH operates on existing coding-agent harnesses, and organizes their executions into iterative planning-coding-testing loops. To sustain improvement across loops, HoH balances repair with capability growth, scopes development into small and verifiable increments, separates implementation-time testing from independent evaluation, and constrains verifiable outputs rather than prescribing agent workflows. It progressively exposes deliverables, role-specific tools, and skills, encourages reuse rather than recreation, and maintains versioned project histories. On GameCraft-Bench, FrontierSWE, and ProgramBench, three harness-model pairs (Codex with GPT-5.5, OpenCode with DeepSeek-V4-Pro, and Pi with MiniMax-M3), HoH consistently outperforms the corresponding standalone harnesses, achieving an average relative gain of 52.25 percent and a maximum gain of 82.86 percent after three iterations. In a multi-day deployment with more than 70 iterations, HoH autonomously develops a first-person-shooter game, featuring a coherent storyline, fully implemented core mechanics, human-playable experience, polished visuals and integrated audio. Github: https://github.com/Flesymeb/HarnessOfHarness Project Page: https://flesymeb.github.io/HarnessOfHarness/
Read the full abstract
High Energy Physics (Phenomenology)
Gravitational Waves from Reheating beyond Instantaneous Thermalization
We study gravitational-wave production during perturbative reheating without assuming instantaneous thermalization of the inflaton decay products. The injected energetic particles thermalize through an in-medium cascade of nearly collinear splittings subject to the Landau-Pomeranchuk-Migdal effect and are ultimately isotropized by elastic scatterings near the thermal scale. The non-thermal hard population present before complete thermalization produces an additional gravitational-wave component through hard-soft scatterings, with $\Omega_{\mathrm{GW}}\propto f^{1/2}$ below the injection-scale turnover. The same isotropization erases the directional information underlying the vacuum $1/k$ bremsstrahlung soft pole, changing the deep-infrared spectrum from $\Omega_{\mathrm{GW}}\propto f$ to $\Omega_{\mathrm{GW}}\propto f^3$.
Read the full abstract
Computer Vision
CameraEditor: Camera-Controlled Image Editing via Video-Prior Sequential Modeling
Beyond semantic content, camera parameters play a pivotal role in dictating the geometric perspective and appearance of any given image. While recent image editing models excel at semantic and stylistic manipulation, they struggle with explicit camera parameter control. When handling large perspective shifts, instruction-driven models face a dilemma: they either suffer from structural tearing or generate conservative outputs that ignore geometric instructions. To address this, we introduce CameraEditor, a framework that reformulates camera-controlled editing from a spatial problem into a temporal sequence prediction task. By leveraging the temporal coherence of video diffusion models, our approach integrates an explicit geometric perception module with a dynamic reference routing mechanism. This allows us to construct geometrically rigorous visual reference pairs via dynamic panorama cropping, overcoming the ambiguity of text-based instructions. Furthermore, CameraEditor strategically inserts intermediate transition frames to decompose large perspective shifts, providing a robust temporal buffer that preserves content identity and spatial coherence. We construct a training dataset of 5,760 instances. As an independent contribution, we introduce CamEditor-Bench, a model-agnostic evaluation suite of 462 test cases. Extensive experiments demonstrate that CameraEditor achieves state-of-the-art camera control precision and source identity preservation, outperforming existing methods.
Read the full abstract
Networking
The Role of Collective Perception and 5G NR-V2X Sidelink in Road Safety
Vehicles and roadside infrastructure are increasingly equipped with sensors capable of perceiving their surroundings. Sharing this information through vehicle-to-everything (V2X) communications is a key enabler of Day-2 applications and is supported by the ETSI collective perception service (CPS). While CPS is expected to play a fundamental role in future intelligent transportation systems, its operation may significantly increase channel load, posing challenges in terms of radio resource utilization, communication reliability, and information management. This paper reviews the current status of CPS standardization and investigates its impact in dense deployment scenarios where connected vehicles communicate through fifth-generation (5G) New Radio-V2X (NR-V2X) sidelink (SL) communications. The main contribution is a realistic evaluation of communication reliability, latency, channel occupancy, and information usefulness under different object-selection strategies for collective perception messages. The analysis is conducted through a network-level simulation framework integrating empirical object traces derived from real-world datasets, thereby avoiding the limitations of synthetic traffic models. Results show that perception message generation and radio access mechanisms are tightly coupled and should be jointly designed to maximize the benefits of collective perception services.
Read the full abstract
Physics Education
What do student responses to the car-truck problems tell us? An investigation into two Force Concept Inventory questions
Concept inventories are widely used in Physics Education Research to uncover student misconceptions and measure the effectiveness of instruction. Two questions in the widely used multiple-choice Force Concept Inventory concern a car that is pushing a truck, as the car is accelerating and then when a steady speed has been reached. We investigate the premise that correct answering of these questions is because of good conceptual understanding of Newton's Third Law, by analyzing responses to a modified inventory that randomly offers students multiple-choice or free-text versions of the questions and follows these with questions in which students select the physical principle(s) they used. The results from 674 attempts from students at six universities show a similar pattern of results for multiple-choice and free-text versions, also comparable with Force Concept Inventory results from one of the universities in the preceding five years, but with marked differences relative to early published Force Concept Inventory data. Free-text responses and the law-selection sub-questions can reveal more about conceptual understanding than can be seen in multiple-choice responses alone. For the accelerating system (Question 15), correct responses are usually attributed to Newton's Third Law and incorrect responses are usually attributed to Newton's Second Law. When speed is constant (Question 16), around 90% of responses are correct, but a significant number of these are attributed to Newton's First Law or the Superposition Principle rather than Newton's Third Law, especially pre-instruction. We conclude that students are confusing balanced forces on a single object moving at constant speed with the correct equal-and-opposite interaction pair between two objects, and correct responses to Question 16 may be concealing an underlying misconception.
Read the full abstract
Superconductivity
Superconducting diode effect from field-induced $s+if$ pairing in Ising superconductors
The in-plane critical field of Ising superconductors exceeds the Pauli limit by an order of magnitude because the Ising spin-orbit coupling locks the electron spins out of the basal plane. The same locking converts an in-plane Zeeman field into a source of equal-spin triplet Cooper pairs, so that the field-driven condensate acquires an $s+if$ character. We show that this conversion channel also generates Lifshitz invariants, the odd-in-momentum terms of the Ginzburg-Landau expansion responsible for the superconducting diode effect. When the basal mirror symmetry of the monolayer is lifted by a substrate or a gate, the field-induced triplets couple linearly to the Cooper-pair momentum, and an intrinsic diode response develops whose strength is set by the ratio of the Zeeman and spin-orbit energies rather than by the small ratio of the spin-orbit and Fermi energies familiar from parity-mixing mechanisms. We construct the symmetry-constrained two-component Ginzburg-Landau theory of the coupled singlet and triplet order parameters and derive all of its coefficients from the microscopic model of an Ising superconductor; the complete functional, including all gradient and quartic terms, is generated by a single pair-breaking function of temperature, field, and Cooper-pair momentum. An attractive triplet channel reshapes the diode response: at weak fields it suppresses the efficiency through destructive interference between the direct and the collective-mode conversion paths, while at strong fields it extends the diode regime well beyond the singlet-only critical field, with the maximal efficiency reached along the triplet-enhanced phase boundary. The diode effect thereby serves as a transport probe of a hidden triplet pairing channel and of the field-induced $s+if$ state.
Read the full abstract
High Energy Physics (Phenomenology)
Inelastic Dark Matter Signature at High Recoil Energy in LUX-ZEPLIN and CRESST
The LUX–ZEPLIN Collaboration recently reported one event at $E_{\mathrm{nr}}=248\pm23~\mathrm{keV}$ with an exposure of $2.84~\mathrm{tonne\cdot yr}$. In this Letter, we interpret this feature using endothermic dark matter (DM). We consider both direct scattering of the surviving ground-state halo component and the contribution of excited states produced by terrestrial upscattering. Our calculations show that explaining the high-energy event requires $m_\chi\gtrsim500~\mathrm{GeV}$ and a mass splitting of $\mathcal{O}(300)~\mathrm{keV}$, for which the production of excited states inside the Earth is kinematically forbidden. For $\bar{\sigma}_n=10^{-37}~\mathrm{cm^2}$, an illustrative two-bin likelihood analysis yields a representative best-fit point at $(m_\chi,\delta)\simeq(1.105~\mathrm{TeV},350~\mathrm{keV})$. The preferred parameter region may be tested by the planned CRESST upgrade.
Read the full abstract
Information Theory
Binary Multiple-Node-Erasure-Correcting Codes over Complete Graphs: Constructions, q-Ary Metric Balls, and Duality
Codes over graphs store symbols on edges of a complete undirected graph on $n$ vertices with a self-loop at each vertex; erasing a node removes all its incident edges. Extending Yohananov, Efron, and Yaakobi's construction, we obtain binary codes correcting any three node erasures at infinitely many prime lengths $n$, unconditionally, with redundancy $3n-2$, one bit above the graph Singleton bound. The earlier construction has the same redundancy but requires $2$ to be primitive modulo $n$, which is not known for infinitely many primes. We give a deterministic decoder for these triple-node codes using $O(n^2)$ binary operations and construct Singleton-optimal binary triple-node codes at $n=6,8,10,12$. When $2$ is primitive modulo an odd prime $n$, we extend their double- and triple-node codes to binary codes correcting $\rho$ node erasures with redundancy $\rho n-(\rho-1)$ for $2\leq\rho\leq(n+1)/2$, saving $\rho-1$ bits over the binary symmetric rank-metric construction when $\rho<(n+1)/2$. Over every finite field of size $q$, the node weight of a word is the least number of vertices whose erasure removes its support. We give exact formulas for ball volumes in this node metric and prove that, for fixed $q,t$, the radius-$t$ volume is asymptotic to $\binom nt q^{tn-\binom t2}$ as $n\to\infty$. We obtain existence and covering bounds; for fixed $q,\rho$, we show that linear $\rho$-node-erasure-correcting codes exist with redundancy at most $\rho\log_q n+O_{q,\rho}(1)$ above the graph Singleton bound. Finally, reformulating Yohananov and Yaakobi's criterion, we show that a linear code over any finite field is Singleton-optimal for $\rho$ node erasures if and only if its dual is Singleton-optimal for erasures of the coordinates inside any $n-\rho$ vertices.
Read the full abstract
Algebraic Topology
Keller sequences of exact $\infty$-categories
We study a family of fibre-cofibre sequences in the $\infty$-category of exact $\infty$-categories, which we call Keller sequences, generalising the notion of Verdier sequences of stable $\infty$-categories. We show that algebraic K-theory is localising with respect to this family, and use this to give a categorical proof of Quillen's resolution theorem. Our arguments hinge upon a mapping space characterisation of special inclusions as studied by Keller, from which we deduce another proof of Keller's theorem that such inclusions induce fully faithful functors on stable envelopes.
Read the full abstract
Computational Geometry
Efficient K-Visibility Query in Polygons
This paper investigates $k$-visibility, where a line of sight can penetrate up to $k$ obstacles. While computing the $k$-visibility polygon from a single query point is well-studied, existing spatial preprocessing approaches rely on full $O(n^2)$ line arrangements through all vertex pairs without characterizing the minimal set of topological boundaries. We present a refined cell decomposition framework that isolates the exact geometric events governing $k$-visibility: primary vertex horizon lines and secondary mutually critical hinge lines. We prove that this minimal set of partition lines yields a spatial decomposition of $\Theta(n^4)$ cells within which the combinatorial structure of the $k$-visibility polygon remains strictly invariant. By leveraging a combinatorial $\delta$-compression scheme across cell boundaries, we achieve an overall storage complexity of $\mathcal{O}(n^4)$ while supporting optimal $\mathcal{O}(\log n + m)$ query time to reconstruct explicit $k$-visibility polygons of size $m$. Our framework naturally extends to polygons containing holes.
Read the full abstract
Plasma Physics
A Magnetothermodynamic Theory of Energy Transport In Collisionless Plasmas
Presented here is an exploratory conceptual framework with the aim of describing electromagnetic and thermal energy transport and transfer in collisionless plasmas in a coherent way, starting from an arrow-of-time perspective. That is, arguments about the transport of energy are fundamentally based on the idea that systems tend to fill the phase space available to them. First, we describe electric fields, magnetic fields, and plasma particles as interacting subsystems that are thermodynamically linked in specific ways. Second, we discuss the importance of the stochastic electric field in the transport of plasma particles down energy density gradients until the energy flux contributions from so-called electrothermodynamic (ETD) interactions become sufficiently uniform. Third, the logic of ETD transport is extended to a magnetized system invoking 2 types of magnetothermodyamic (MTD) transport: 1) transport of electromagnetic energy through stochastic Poynting flux and 2) plasma energy transport through stochastic drift energy fluxes associated with local ExB drift motion. The magnitudes of the ETD and MTD mechanisms are tied to local characteristic scales of the electric field fluctuations and critical scales are derived beyond which ETD mechanisms dominate MTD mechanisms. Finally, we discuss the interplay between electrothermodynamic and magnetothermodynamic equilibria (ETE and MTE), dynamics in the small critical scale (MTD) limit, and the potential relevance of this type of description to the problem of reconnection onset and scale coupling.
Read the full abstract
Computer Vision
RadMatch: Auditable Radiology Report Evaluation via Finding-Level Matching
As AI systems are increasingly used to draft radiology reports, reliably evaluating their clinical quality remains a critical challenge. Large language model (LLM)-based metrics are now the best-correlated with radiologist judgment, yet they output a single opaque score that neither a clinician nor a model builder can easily interpret or audit. We introduce RadMatch, a multi-stage, LLM-based metric that decomposes report comparison into a structured finding-level matching with significance-aware scoring and error characterization across seven clinical attribute dimensions (status, location, severity, morphology, certainty, longitudinal comparison, and measurement). The main score is the actionable-error count, both interpretable and auditable. Candidate findings are graded correct, partial, or incorrect, and unmatched findings are counted as missed or hallucinated. Triage and actionable safety recall/precision and per-subset views add complementary, deployment-oriented lenses. Across two expert benchmarks, RadMatch is the most clinically aligned metric, matching inter-radiologist agreement on ReXVal and more than doubling the best prior metric on the harder RadEvalExpert. Relying only on few-shot prompting, it is designed to extend to other modalities and anatomies. We will release RadMatch as open-source code with an interactive dashboard for inspecting results.
Read the full abstract
Computational Complexity
SVP Is NP-Hard for Some Rank-2 Cyclotomic Modules
Let $q$ be an odd prime. Let $\zeta_q$ be a primitive $q$th root of unity, and put $K=\mathbb{Q}(\zeta_q)$, with ring of integers $\mathcal{O}_K=\mathbb{Z}[\zeta_q]$. As $q$ ranges over odd primes, we prove that the decision version of the Shortest Vector Problem ($\mathsf{SVP}$) in the canonical $\ell_2$-norm is $\mathsf{NP}$-complete on full-rank free submodules of $\mathcal{O}_K^2$ by giving a deterministic polynomial-time many-one reduction from Exact Cover by 3-Sets (X3C). The module rank is fixed at two. The main obstacle is closure under the action of $\mathcal{O}_K$. A module containing a nonzero vector also contains every scalar multiple of that vector by a nonzero element of $\mathcal{O}_K$, and some of these multiples may be shorter. Three ideas overcome this obstacle. First, we map the Bennett–Peikert Reed–Solomon lattice to a principal cyclotomic ideal and use Wan's point-count estimates to prove that a coset of this ideal contains many binary coefficient representatives. Second, we construct a checker whose canonical squared norm encodes the X3C equations, with a controlled nonnegative error term. Third, the checker and a second module coordinate combine with a separation bound for ideal cosets to rule out short vectors in NO instances. Each constructed instance consists of an odd prime $q$, two integral generators whose $2\times2$ generator matrix has nonzero determinant, and an integer squared threshold. The construction also gives $\mathsf{NP}$-hardness of search-$\mathsf{SVP}$ under deterministic polynomial-time Turing reductions with at most one oracle call.
Read the full abstract
Theoretical Economics
Freemium Model for Information Provision
The paper explores a theoretical freemium model for the sale of information, drawing on mathematical tools used in the study of repeated zero-sum games and Bayesian persuasion. Unlike standard Bayesian persuasion models, the information seller (IS) is indifferent to the actions taken by the information buyer (IB) and is concerned solely with maximizing the revenue from selling information. Offering some information for free may increase the IB's willingness to pay for additional information. The information that the IB seeks is about the state of the world. Initially, the IB only knows the prior distribution over possible states. The IS supplies both free and paid information through signals whose state-dependent distributions determine the IB's posterior via Bayes' rule. The IB's utility is a function of the posterior. An optimal free signal is one that maximizes the IS's expected revenue from the subsequent paid signal. That revenue is equal to the IB's expected utility gain when moving from the posterior induced by the free signal to that induced by the paid signal. The paper characterizes the optimal free and paid signals and derives a formula for the maximal revenue in terms of the IB's utility function. It shows that a revenue gain for the IS from the provision of free information is accompanied by a loss to the IB. Whether free information can increase the IS's revenue depends on the form of the IB's utility function. In the two-state case, that dependence is fully characterized. In the general case, only necessary conditions are obtained. In particular, if the IB's utility function is convex, the IS can never profit from providing free information. This occurs, in particular, when the IB uses the information to solve a decision problem. By contrast, when the IB is engaged in a strategic interaction with a third party, the IS may benefit from providing free information.
Read the full abstract
Econometrics
Which Policy Works, and Where? Estimation and Inference for State-Level Treatment Effects in Difference-in-Differences
Policies with a common objective and implementation date may differ in details or context. We distinguish the aggregate average treatment effect on the treated (ATT) from sub-aggregate ATTs defined by implementation cohort, jurisdiction, period, or policy type. UN-DID and DID-INT, two DiD estimators that construct jurisdiction-by-time effects, estimate these ATTs under parallel-trends conditions matched to the aggregation. In CPS placebo-law simulations, randomization inference is generally well-sized, though some jurisdiction-specific tests are conservative. The jackknife can be undefined for sub-aggregate ATTs; when defined, it over-rejects with few treated or comparison jurisdictions. Estimands and inference methods should match the policy question and implementation setting.
Read the full abstract
Artificial Intelligence
Parsing the Stream: A Live Trace Model for Long-Horizon Agents and Their Observers
A long-horizon agent's trace outgrows both of its consumers: the human observer monitoring the run, and the agent itself, whose bounded context the trace must be folded back into. We present a live trace model, an append-only event ledger folded incrementally into typed run state and compiled into per-consumer views, and evaluate it for both consumers against deterministic ground truth. For the observer side, evaluated with an LLM reader as proxy, the compiled view answers monitoring questions using approximately 14x and 15x fewer input tokens (by reader) and at 5-7x lower cost than a budget-capped single-call reading of the raw trace, with higher accuracy (0.85-0.87 versus 0.48). Because the questions were co-designed with the view schema, we treat the token and cost reduction, conditional on schema coverage, as the transferable result. For the agent, on 120-link sequential-dependency tasks, mechanisms that maintain the task's running statistic in per-step state succeed where full-context prompting fails (30/30 versus 8/30 under a clean protocol, n=30, labeled descriptive owing to benchmark-system co-development); a prompt-level scratchpad matches the fold's accuracy at lower cost, and a two-arm decomposition attributes the fold's accuracy to its deterministic aggregate and its cost advantage to its compactness. The fold's remaining value over cheaper alternatives is deterministic auditability and serving the observer from the same state. We derive eleven candidate requirements for trace folding from observed failures and delimit them with an order-sensitive task family on which the fold ceases to help. Code, benchmarks, a regenerable synthetic corpus, and all workbench traces are released.
Read the full abstract
Networking
Mobile Backscatter Communication for the Battery-less Internet of Things
We enable backscatter communication in the battery-less mobile Internet of Things (IoT). Backscatter communication is extensively studied in static settings. Existing designs are, however, fundamentally mismatched with mobility and time-varying energy patterns. Channel conditions rapidly fluctuate, impacting the achievable data rates and thus transmission costs. Energy availability varies unpredictably, possibly forcing devices to remain quiescent to recharge energy buffers. The two issues compound each other: while recharging, a battery-less mobile IoT device may miss more favorable channel conditions. We design a lightweight decision system that dynamically determines when to transmit by checking short-term trends in signal strength, while using Non-volatile Memory (NVM) to retain packets in unfavorable channel conditions and across energy failures. Using a prototype we built and real-world mobility and power traces, we compare our design against a rate-adaptive baseline that only considers the instantaneous channel conditions. Experimental results show that our system improves throughput by up to 5.16x while reducing transmission energy consumption by up to 47.3%, with only 0.23% - 7.3% additional energy overhead.
Read the full abstract
Differential Geometry
Removability of non-isolated singularities for Einstein metrics and RCD spaces
In this paper we establish removable singularities results for Einstein metrics and for metrics with Ricci curvature bounded below. Let $n\geq 2$. On a closed $n$-manifold, we show that an $L^\infty$-Riemannian metric whose Ricci curvature is bounded below outside a singular set of codimension $> 3- \frac{1}{n-1}$ canonically extends to an $\mathrm{RCD}$ space. As a consequence, using a new removable singularity theorem for Einstein metrics, we prove that in dimension $4$ any Einstein metric with $L^\infty$ singularities of codimension $>3-\frac{1}{3}$ extends smoothly across the singular set, possibly after changing the smooth structure. In higher dimensions, we construct a $C^{1,\alpha}$-Riemannian manifold structure on the regular set of a non-collapsed $\mathrm{RCD}$ space that is a Riemannian manifold with bounded $|\mathrm{Ric}|$ outside a set of codimension $>2$. Our results can be used to give a proof of Schoen's conjecture on scalar curvature singularities for metrics that are either continuous, or $L^\infty$ and sufficiently close to a smooth background metric.