Number Theory
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Sunday 8 November
We prove the étale-de Rham direction of the $C_{\st}$-conjecture for compactly supported cohomology of smooth partially proper rigid analytic varieties over local fields of mixed characteristic. More precisely, we recover filtered de Rham cohomology and (under a finite-rank assumption) Hyodo–Kato cohomology from compactly supported $p$-adic pro-étale cohomology as derived Hom (in a category of Arithmetic Topological Vector Spaces) with values in $\B_{\pdr}$ and $\B_{\pFF}$ respectively (period rings related to $\B_{\dr}$ and $\B_{\st}$ whose higher Galois cohomology vanishes). The proof uses, via the geometrized basic comparison theorem, Poincaré duality for Hyodo–Kato and (filtered) de Rham cohomology, explicit extension calculations for period sheaves, and fully-faithfulness of the passage from Galois equivariant perfect complexes on the Fargues–Fontaine curve to Arithmetic Topological Vector Spaces.
Artificial Intelligence
FinCutBench: A Continually Updated Benchmark for Measuring the Temporal Boundaries of LLM Financial Knowledge
In financial backtesting and other point-in-time tasks, external inputs can be restricted by date, but model parameters may retain financial facts disclosed later, introducing parametric look-ahead bias. We introduce FinCutBench, a continually updated monthly financial knowledge benchmark for measuring how far a fixed model's financial knowledge extends over calendar time. The current dataset contains 9,000 Chinese and English four-option multiple-choice questions from January 2023 to September 2026, covering Chinese and U.S. markets and four categories of financial information. We construct monthly knowledge profiles using the approximate months in which the tested facts became public. After a profile passes the decline test and knot checks, we define the FinCut Boundary as the first month when the fitted profile reaches its lower level, and report the stability of the estimated date under resampling. Boundary dates and localization precision vary across models. Using the independently measured Transition Start and Boundary as fixed references for downstream sample grouping, we find significant relative declines in U.S. asset-ranking performance and in earnings recall for Sol and Luna. Downstream effects vary across markets and tasks. The FinCut Boundary thus provides an empirical reference for the temporal scope of parametric knowledge in point-in-time financial evaluation, supporting evaluation-window selection and result interpretation.
Read the full abstract
Cosmology and Nongalactic Astrophysics
Detecting Baryon Acoustic Oscillations from Galaxy Clusters: Distance Constraints
We present a measurement of the baryon acoustic oscillation (BAO) feature using the largest galaxy cluster sample with spectroscopic redshift information to date, comprising 169,679 clusters selected from the WH24 catalogue based on the DESI Legacy Imaging Surveys. We measure the two-point correlation function in two redshift bins, $0.17\leq z<0.35$ and $0.35\leq z<0.5$, corresponding to effective redshifts of $z_{\rm eff}=0.25$ and $0.43$, with the BAO feature detected at approximately the $4\sigma$ level or above in both bins. By fitting the measured correlation functions, we obtain dilation parameters of $\alpha=0.97\pm0.07$ and $\alpha=1.01\pm0.04$, corresponding to volume-averaged distance measurements of $D_V/r_d=6.70\pm0.48$ and $D_V/r_d=11.25\pm0.47$, respectively. Using the Planck 2018 sound horizon calibration, we obtain $H_0=66.6 \pm 4.9 \,\mathrm{km\,s^{-1}\,Mpc^{-1}}$ within the flat $\Lambda$CDM framework. We further combine our cluster BAO measurements with cosmic microwave background and Pantheon+ Type Ia supernova observations to constrain dynamical dark energy within the Chevallier–Polarski–Linder parametrization. The resulting constraints are consistent with $\Lambda$CDM, which lies within the $68\%$ joint credible region in the $w_0$–$w_a$ plane. Our results demonstrate the potential of galaxy clusters as complementary tracers for BAO measurements and cosmological constraints.
Read the full abstract
Computer Vision
FastGRT: Accelerating Training of 3D Gaussian Ray Tracing
Ray Tracing enables rendering capabilities that are difficult to achieve with rasterization-based approaches such as 3D Gaussian Splatting (3DGS). However, the training cost of Gaussian-based ray tracing (GRT) is much higher than that of 3DGS. In this paper, we study how ray tracing using Gaussians executes on GPUs and show that a considerable fraction of the training cost arises from how the computation is organized. We develop FastGRT, which decouples ray traversal from blending and reuses work done in the forward pass for the backward pass. FastGRT further incorporates optimizations such as fast gradient accumulation and a hybrid BVH update strategy to tackle other major bottlenecks in the training pipeline. We implement FastGRT in Vulkan. To the best of our knowledge, it is the first Vulkan-based GRT implementation that supports both training and rendering. Compared to 3DGRT, FastGRT improves training time by 1.9$\times$ to 4.97$\times$, while preserving the original rendering formulation. Our code is open source.
Read the full abstract
Optimization and Control
Probabilistic Weapon-Target Allocation Models for the Critical Infrastructure Protection
This paper analyzes methods for solving optimal resource/weapon allocation problems for the protection of critical infrastructure. It examines nonlinear discrete "Attacker-Defender" and "Weapon-Target Allocation" models, incorporating criteria based on the difference in expected losses and risk-related constraints. Methods are proposed for solving these problems by reformulating them as mixed-integer linear programming problems.
Read the full abstract
General Relativity and Quantum Cosmology
Quasinormal modes of a scale-dependent black hole in 5D spacetime with cosmological constant
In this paper, we investigate the quasinormal modes (QNMs) of a scale-dependent black hole in five dimensions with a positive cosmological constant, considering i) scalar, ii) electromagnetic, and iii) Dirac perturbations. After deriving exact analytical solutions within the framework of scale-dependent gravity in five dimensions and selecting a different family of solutions (cf. \cite{Contreras:2019fwu}), we analyze the corresponding effective potentials (given by the lapse function) for each perturbation type. We employ a first-order approximation for the lapse function, as it closely matches the exact case for small $\varepsilon$. Since quantum corrections are expected to be small, our focus is on a relatively low range of the scale-dependent parameter $\varepsilon$. We confirm that the WKB semi-analytical approach can be used to compute the QNMs for these perturbations to the background geometry. We calculate the QNM frequencies for the classical case (for comparison) and for cases where the scale-dependent parameter $\varepsilon$, which quantifies deviations from General Relativity, is non-zero. The results are summarized in a series of tables and figures. Finally, we find that, regardless of the perturbation type, the solutions are stable, as the imaginary part of each QNM frequency is negative.
Read the full abstract
Computer Vision
NightSteps: A Night-Time Dataset and Benchmark for Pedestrian Crossing Intention Prediction
Autonomous vehicles use pedestrian crossing intention prediction (PCIP) to participate safely in traffic. Existing models are largely trained under favourable daytime conditions with sufficient daylight. Little is known about the robustness of these models under night-time and low-light conditions, due to the lack of datasets containing night-time footage. This paper investigates how a dataset, named NightSteps, can be constructed and evaluated as a baseline benchmark. The dataset is based on nighttime dashcam footage collected from six different cities around the world, capturing a wide variety of road conditions. The dataset consists of dashcam-recorded point-of-view footage, raw data of pedestrians' locations with corresponding 'track ids' and ground-truth labels containing accurate information about pedestrian crossing behaviour, resulting in a diverse dataset. Gemini, ChatGPT, and human assessment are used to demonstrate the usability of the dataset by predicting crossings in clips from the dataset and comparing the results with the initial ground-truth labels. These results show that human input achieves an accuracy of 88.17% and an F1-score of 80.21%. Predictions from Gemini and ChatGPT perform lower, with accuracies of 60% and 79.13% respectively, and F1-scores of 42.50% and 63.64%. The evaluation demonstrates that the NightSteps dataset can be used for future model evaluation to test the performance of PCIP, with the aim of improving road safety at night.
Read the full abstract
Statistics Theory
Robust Gaussian process regression: posterior contraction
Gaussian processes form a popular and flexible class of prior distributions with principled uncertainty quantification, but the Gaussian likelihood on which the corresponding posterior relies makes them highly sensitive to outliers and contaminated observations. We study Gaussian process regression when among the $n$ observations, $m$ of these are possibly contaminated. A generic framework to obtain posterior contraction rates in this setting is considered, from a frequentist perspective as the sample size $n$ and the number of corrupted observations $m$ grow. We adopt a generalized Bayes approach in which the log–likelihood is replaced by a robustified version; we discuss various choices thereof, such as one recently proposed by Altamirano et al. that preserves conjugacy. We investigate the minimax rate of estimation under adversarial contamination in terms of the Sobolev smoothness $\beta$ of the regression function and $m$. The question of building adaptive procedures to both $\beta$ and $m$ is investigated, and we show that a polynomial loss for adaptation must be paid in general. We instantiate our results for Matérn Gaussian process and deep ReLU network priors. Numerical experiments for existing and new robust algorithms are provided, illustrating the practical applicability of our results.
Read the full abstract
Artificial Intelligence
Same Error, Different Recovery: How Delivery Changes LLM Agent Behavior
A tool-using language model must decide whether to retry a failed call, correct it, or stop and tell the user. The error text alone may not be enough: the same feedback can lead to different actions depending on where it appears in the conversation. We isolate this delivery problem by varying placement, guidance, retained history and turn structure. Two studies with plans fixed before data collection, across 13 open models from 3B to 70B parameters, compare the next action and subsequent task completion, with exact deterministic reruns. A new user turn that retains the failed exchange avoids the severe stopping failure observed in Llama models when that exchange is removed. It is non-inferior to history rewriting in pooled synthetic comparisons, but is not best for every model. We implement this form as an opt-in recovery layer within a deterministic inference engine and evaluate it on public benchmarks and a coding agent. On ToolMaze, it raises success on explicit errors by 14.2 points with 7-8B models and 10.5 points with 32-70B models; Loud/Quiet, 24 multi-turn tasks with visible failures and silent corruption, scored by final-state agreement, shows no measurable gain. A client given the exact fault kind performs better still. The findings support delivery as a distinct factor in recovery, not an engine-exclusive solution or a universal retry policy.
Read the full abstract
Robotics
Qubit-to-Motion: Traceable Action Provenance in Hybrid QRL for Robotic Bin-Packing Sequencing
Hybrid quantum reinforcement learning has been explored for robotic decision-making, but evaluations often emphasize task success, return, or convergence, leaving unresolved whether quantum circuit outputs influence constrained actions and remain traceable through robot execution. We present Qubit-to-Motion, an evidence-centered framework for support-constrained sequencing and simulated robotic placement. Four Pauli-$Z$ readouts provide the sole learned input to the actor and critic, while a support mask enforces admissible object selection. Each decision is recorded and linked via step indices and SHA-256 fingerprints to a sealed route, deterministic Cartesian and inverse-kinematics waypoints, prescribed joint motion, and MuJoCo contact replay. A frozen checkpoint is evaluated across five deterministic 30-object scene variants. All 150 placements satisfy support, containment, and settling criteria. A fixed-state zero-readout intervention changes $90.0\pm0.0\%$ of decisions, demonstrating dependence on the joint circuit representation. Mean macro travel is $34.537\pm0.206$~m, inference requires $8.60\pm1.27$~ms per route, and replay achieves $8.764\pm0.220~\mu$m mean final position error with 0.356~mm maximum interpenetration. These results establish reproducible decision-to-contact traceability. The implementation is available at \href{https://anonymous.4open.science/r/Qubit-to-Motion-Quantum-Reinforcement-Learning-F7DC}{\textcolor{linkpink}{the anonymous repository}}.
Read the full abstract
Signal Processing
Descriptor: UWB Anchor Self-Calibration Dataset (UWB-ASC)
Anchor self-calibration estimates the positions of fixed nodes of an indoor radio positioning or localization system from observations among the nodes and to mobile devices. It replaces the survey of every anchor, which is costly and goes stale when anchors move. The UWB Anchor Self-Calibration Dataset (UWB-ASC) is a testing ground for such methods, recorded in one industrial hall with ultra-wideband (UWB) two-way ranging. It has three parts. The first is two measured deployments. In the line-of-sight (LOS) deployment, 10 anchors range with a walking tag tracked by motion capture. In the non-line-of-sight (NLOS) deployment, 12 nodes range among obstructions, and one of them roves over surveyed markers in 20 quasi-static sessions. Both deployments also hold static sessions, and a total station surveyed every fixed node. The second part is a seeded range simulator that draws synthetic ranges on the surveyed geometry with adjustable error components, from noise-free ranges to NLOS errors, for benchmarks at controlled error levels. Its 81 frozen draws in 5 conditions serve as the reference. The third part is a set of readers that convert the ranges among the static nodes of two public UWB datasets, TWR-CLOUD-Bern and selfloc, into the same format without redistribution. All parts share open Parquet and YAML files, in which topology views select the ranges for each class of calibration method. Raw recordings with rebuild scripts, quality reports, and field labels of the line of sight complete the dataset.
Read the full abstract
Optimization and Control
Recursive Pricing and Incumbent Subtree Column Generation for Solving Optimal Classification Trees
Solving optimal classification tree (OCT) problems is computationally challenging because branching, routing, and leaf-assignment decisions must be optimized jointly. We develop an incumbent subtree column generation algorithm (ISCG-OCT) that exploits the recursive structure of classification trees. We first introduce a mixed set-packing and set-partitioning formulation (MSSF). We then construct a restricted master problem whose columns represent feasible subtree patterns. Unlike conventional dual-based column generation, our incumbent-centered pricing evaluates objective improvements relative to the current solution. We establish a recursive decomposition in which fixing a branching decision separates the remaining pricing problem into independent subtree subproblems, enabling exact dynamic programming. We also develop threshold-based variants that trade optimality guarantees for computational efficiency. Computational experiments on 27 benchmark datasets demonstrate competitive training and testing accuracy and substantial reductions in solution time on many instances. When initialized with a CART solution, ISCG-OCT can certify global optimality faster than warm-started mixed-integer programming solvers on selected instances. These results demonstrate how recursive pricing and structured subtree optimization can improve the computational efficiency of exact classification tree methods.
Read the full abstract
Metric Geometry
No regular icosahedron has a rational point on every face: a proof checked in Lean
We prove that no regular icosahedron in $\mathbb{R}^3$ has a point with rational coordinates on each of its twenty closed faces, in every position, size and orientation, reflections included. This answers a question of I. Bogdanov on MathOverflow. A field automorphism of $\mathbb{C}$ that sends $\sqrt5$ to $-\sqrt5$ reduces the problem to the following statement: for every complex orthogonal $3\times3$ matrix $T$ some real linear combination of $\mathrm{Re}\,T$ and $\mathrm{Im}\,T$ separates ten intervals determined by the faces. We prove this statement by a computer-assisted case analysis over rational parametrizations of rotation frames, in which every inequality is an exact polynomial bound over $\mathbb{Z}[\varphi]$, $\varphi=(1+\sqrt5)/2$; no floating-point arithmetic and no Lipschitz estimates are used. The whole argument, including the $48$ certificates, is formalized in the Lean 4 proof assistant with Mathlib and checked by the Lean kernel; the main theorem depends only on the standard axioms. The mathematical argument and the formalization were produced by AI systems (OpenAI Codex and Anthropic Claude) under the direction of the author; see the statement at the end of the paper.
Read the full abstract
Computation and Language
Continuous modeling of manipulative participation for stance analysis: Online discourse on the 2024 Street Animal Law in Türkiye
Studies of online political discourse commonly infer public opinion through stance analysis while implicitly assuming that observed user participation is authentic. Meanwhile, studies focusing on detecting manipulative participation are typically modeling it as a binary phenomenon, overlooking the varying degrees to which users may engage in coordinated or inauthentic behavior. This study proposes a continuous modeling approach for manipulative participation and examines its integration with stance analysis. Using online discourse on the 2024 Street Animal Law in Türkiye as a case study, we collected a dataset comprising 7,320,574 tweets from 537,994 users. From this corpus, 301 users were selected through a high-information sampling strategy and manually annotated by domain experts using a 10-criterion rubric that yields user-level continuous manipulative participation scores. Evaluation under a unified user-level cross-validation protocol showed that the proposed framework accurately estimates both manipulative participation and user stance, with gradient boosting methods providing the best manipulative participation estimates and BERTurk achieving the highest stance classification performance. External validation on an independently annotated user set demonstrated good generalization performance. The findings support modeling manipulative participation as a continuous construct rather than a binary label, providing a more informative basis for computational analyses of the relationship between manipulative participation and stance in online political discourse.
Read the full abstract
Representation Theory
Representations of the Weyl algebra modulo $p^2$
Let $\mathbf{k}$ be an algebraically closed field of characteristic $p>0$. Let $R=W_2(\mathbf{k})$ be the ring of Witt vectors of length $2$. In this note, we are concerned with the study of representations of the first Weyl algebra that are finite-rank free $R$-modules ($R$-lattices). Our main result shows that given an $R[x]$-lattice $V$, it admits a left module structure over the Weyl algebra (equivalently, an $R$-linear connection) if and only if all of its indecomposable summands have the form $R[x]\otimes_{R[x^p]}N$, where $N$ is an indecomposable $R[x^p]$-lattice of rank divisible by $p$. We also classify, up to isomorphism, all lattices over the $p$-Weyl algebra (defined by $[y,x]=p$) for which the action of $x$ modulo $p$ is regular nilpotent.
Read the full abstract
Number Theory
On the $\Delta$-primitive squares conjecture
A natural number $n$ has the $\Delta$ property if some difference of complementary divisors of $n$ is the sum of two such differences; equivalently, if some split graph has a factor graph containing an $n$-simple triangle. In additive terms, $n$ has the $\Delta$ property exactly when the positive part of the factor-difference set of $n$, introduced by Erdős and Rosenfeld, is not sum-free. The set $\mathbb{N}(\Delta)$ of such numbers is closed under multiplication by squares, and it was conjectured that it contains infinitely many $\Delta$-primitive perfect squares. We prove this conjecture, and show that these squares can be chosen to avoid any prescribed finite set of primes but $3$.
Read the full abstract
Computation and Language
Beyond the Transcript: Measuring and Improving the Use of Acoustic Information in Audio Language Models
Audio language model benchmarks provide comprehensive evaluation across diverse tasks. However, some tasks can be answered from the transcript or even the question alone, and benchmark scores can therefore confound acoustic understanding with language modeling ability. We distinguish lexical tasks, for which the transcript usually suffices, from global-attribute and within-recording-structure tasks, which require the model to analyse the recording beyond its words, either as a whole, for properties such as the speaker and emotion, or within it, for patterns such as stress and intonation. We evaluated 19 open models and Gemini 3.1 Pro on 92 tasks from 6 benchmarks and estimated the recording gain, the accuracy advantage of the original recording over references that retain its words, by comparing the recording with synthetic readings of its transcript and by adding it to the transcript. On attribute and structure tasks, the open models fall behind Gemini by 28.6 and 41.9 percentage points, and about half or more of each deficit lies in the recording gain. Nevertheless, linear probes in Gemma-4 12B and Qwen2.5-Omni-7B show that the acoustic information needed for answering is already readable in early language-model layers, while the readout of many cues declines in the layers that follow. We therefore constructed training data and trained LoRA adapters in these layers, with a content anchor, which raises attribute and structure accuracy by 12.8 to 44.4 points in the two models and keeps mean lexical accuracy within 1 point of the base model. Most of the improvement appears as larger recording gains. These results suggest that training improves the use of acoustic information that the models already encode.
Read the full abstract
Computational Complexity
Closing the Complexity Gap for Wheeler Graph Recognition on $d$-NFAs
Wheeler graph recognition asks whether a directed graph whose arcs are labeled by symbols from a totally ordered alphabet admits a Wheeler order, a total order on its vertices compatible with the labels of arcs and the order of their tails. A $d$-NFA is a nondeterministic finite automaton in which at most $d$ transitions with any given label leave each state. For Wheeler graph recognition on $d$-NFAs, Alanko et al. gave a polynomial-time algorithm for $d \leq 2$, whereas Gibney and Thankachan proved NP-completeness for every fixed $d \geq 5$. The computational complexity for $d = 3$ and $d = 4$ remained open. We resolve both open cases by extending the reduction of Gibney and Thankachan with a new encoding of clauses of Not-All-Equal 3-SAT. Specifically, we prove that recognition is NP-complete even for directed acyclic graphs over an ordered binary alphabet with a unique source reaching every vertex and total out-degree at most three at every vertex. Consequently, Wheeler graph recognition on $d$-NFAs is NP-complete for every fixed $d \geq 3$.
Read the full abstract
Cryptography and Security
Treating the Model as an Untrusted Component: Deterministic Policy Enforcement, Human Approval and Taint Tracking Against Prompt Injection in Tool-Using Agents, with Validation on Open-Weights Models
Tool-using LLM agents act on instructions embedded in the data they read, so prompt injection is a structural rather than a prompt-wording problem. We compare in-prompt defences (hardened prompts, spotlighting, LLM detectors) with deterministic enforcement outside the model (four reference monitors of increasing strength), human approval and taint tracking, judged by one effect-based oracle independent of every defence in an executable environment of 13 tasks and 12 tools, with argument-fuzzing attackers and severity-weighted harm. In an evaluation with real open-weights models (Qwen2.5-1.5B and Qwen2.5-7B; 156 attack episodes per defence and model), the models proposed the attacker's call in 64% and 84% of attack episodes behind the task-bound monitor, which admitted 1.3% and 7.7%; guardrail effects depended on the model (a prompt guardrail moved obedience from 67% to 65% and from 86% to 55%, spotlighting plus a detector to 12% and 0%) against static attacks. Simulation with an assumed model covers what no real run did: under the assumed erosion law a persistent adaptive attacker or a compromised model drives every guardrail stack to 100.0% attack success, whereas the task-bound monitor holds severe actions at 0.0%; its residual (harm 0.69/10) falls by 62% under taint-triggered human review, and a human layer without a deterministic monitor fails against approval spamming unless each episode has a deterministic approval budget. Implementation quality matters: a naive monitor admits 18 of 33 effective attack variants, a hardened one 6 and a capability-bound one 1. The real-model evidence is limited to two open-weights models and static attacks, the human-approval and taint results are simulation-only, and the larger model's detector stack contradicts the assumed erosion law. Code, data, a two-rater oracle audit and a notebook for repeating the real-model evaluation are provided.
Read the full abstract
High Energy Physics (Phenomenology)
HEART: a resonance-cascade model of soft hadron production in p+p collisions at 158 GeV/c
Fischer and co-workers interpret the NA49 identified-hadron spectra in p+p collisions at 158 GeV/c as the sum of a target and a projectile component, each the decay of an excited hadronic system. This paper supplies an event-level realization of that picture, the HEART model (Hadronic Excitations And Resonance Transitions), and tests it on the published NA49 data. Single diffraction, double diffraction, and non-diffractive exchange produce color-singlet excited systems. Each system is a catalogued resonance below a duality boundary of 1.7 to 2.2 GeV or a member of a resonant continuum above it. It de-excites by successive two-body transitions with exact conservation laws, and one-pion exchange produces the leading neutrons. Fourteen parameters are fitted to 1,254 points of the $\pi^\pm$, $K^\pm$, $p$, and $\bar{p}$ double-differential cross sections and eight neutron points. Under an error model that profiles the correlated systematic uncertainties, with a declared 5% model floor, the fit yields $\chi^2 = 2263$ on 1248 degrees of freedom. On the same points, EPOS-LHC, UrQMD, and PYTHIA 8 at their published tunes give $\chi^2$ values 4.2 to 14.4 times higher. The model needs no central source. The forward $K^+$ deficit, the forward $K^*(892)^0$ excess, and the flat $\Lambda$ $x_F$ distribution point to the momentum sharing at the vertex where the leading baryon turns into a hyperon. The $\Xi^-$ reaches 0.61 of the NA61/SHINE yield. The charged multiplicity of a heavy excited system grows linearly with its mass, faster than in $e^+e^-$ annihilation. The observables that distinguish a cascade from string and fireball models lie in the fragmentation region, where wide-acceptance fixed-target spectrometers can measure them.
Read the full abstract
High Energy Physics (Phenomenology)
Hadronic final-state effects on open-charm production at LHC energies
We investigate the influence of hadronic final-state interactions on open-charm observables in pp and Pb+Pb collisions at LHC energies within the microscopic Parton-Hadron-String Dynamics (PHSD) transport approach. Calculations with and without charm-hadron rescattering are compared with ALICE measurements and predictions of the Statistical Hadronization Model for charm. The relative integrated $D$-meson multiplicities are essentially insensitive to hadronic interactions. PHSD reproduces the weak system dependence of the nonstrange $D$-meson ratios and the increase of $D_s^+/D^0$ from pp to Pb+Pb collisions, consistent with the statistical-model trend from strangeness undersaturation towards equilibrium. In contrast, hadronic rescattering redistributes the transverse momentum spectra at low and intermediate momenta and enhances the elliptic flow. These results demonstrate the importance of accounting for the hadronic evolution when extracting QGP properties from differential open-charm observables.
Read the full abstract
High Energy Physics (Phenomenology)
Spin entanglement in $pp$ and $pA$ collisions: high-energy TMD effects and fragmentation probes
Spin-spin correlations of a quark-antiquark pair produced in proton-proton or proton-Ion collisions are investigated in $k_T$-factorization within the shockwave formalism. We construct the full spin-density matrix of both inclusive and diffractive channels in the limit in which the two quarks are back-to-back. In the inclusive channel, due to linearly polarized gluons, we show that a $\cos (2 \phi)$-modulation of the concurrence, stabilizer Rényi entropy, and quantum discord emerges and depends on the interplay between six small-$x$ gluon TMDs. The modulation is predicted to be maximally strong and independent of the transverse momentum imbalance, $q$, from a dilute (BFKL) perspective, while it is considerably suppressed by saturation effects. In diffraction, we observe that the pair is always entangled and exhibits the same properties found in diffractive DIS. We explain how these properties follow directly from the exclusive and eikonal nature of the Pomeron exchange. We trace this behavior to a spin-density matrix of rank at most two with maximally mixed marginals. For this class of states, entanglement and Bell nonlocality coincide even when the state is mixed, while the explicit form of the light-front helicity amplitudes ensures that both persist in the whole phase space. We suggest accessing the spin correlations in unpolarized hadron collisions through the Collins and dihadron fragmentation functions. We point out that, in inclusive production, for massless quarks, suitably chosen double-Collins asymmetries provide promising entanglement witnesses.
Read the full abstract
Distributed Computing
Verifiable, Privacy-Preserving and Collaborative Multi-Objective Optimization with NSGA-II on Blockchain
Collaborative multi-objective optimization allows independent problem solvers to contribute their solutions to a shared problem. However, their validity cannot be reliably verified in open settings if the parties are reluctant to disclose their solutions. Therefore, this paper proposes a novel zk-NSGA-II protocol for verifiable, privacy-preserving and collaborative multi-objective optimization on blockchain. To the best of our knowledge, this is the first work that executes Non-Dominated Sorting Genetic Algorithm II (NSGA-II) on the RISC Zero Zero-Knowledge Virtual Machine (zkVM) where evolutionary generations are certified by succinct proofs that are verified on-chain. Problem solvers maintain their own independent populations as the set of solutions in the decision space and their corresponding objective vectors in the objective space. They contribute to a global on-chain Pareto front by aggregating their local off-chain Pareto fronts according to the rule of non-domination which admits a contribution if it is non-dominated with respect to the existing contributions in the global front. To prevent fabricated contributions, population states across generations are linked through hash chains. The correctness of the zk-NSGA-II protocol is first evaluated on twelve problems from the ZDT and DTLZ benchmarks. Second, its feasibility is shown in a real-world portfolio optimization case study with the Hang Seng dataset. Performance on these problems is measured with nine different metrics across three complementary perspectives for optimization, zero-knowledge proof and blockchain. The empirical findings indicate that trustless collaborative multi-objective optimization is feasible with the existing zero-knowledge and blockchain technologies.
Read the full abstract
Number Theory
A rank-zero $2$-converse to the Gross–Zagier and Kolyvagin theorem for elliptic curves with abelian two-division field
Let $E/\mathbb{Q}$ be an elliptic curve with abelian two-division field. We prove that finiteness of ${\mathrm Sel}_{2^\infty}(E/\mathbb{Q})$ implies $L(E,1)\ne0$. The proof follows the quadratic twist deformation approach of OpenAI: Beilinson–Kato elements are interpolated over the cyclotomic $\mathbb{Z}_2$-extension and over families of quadratic twists, and congruences in the resulting group ring transfer non-vanishing from the twists to $E$. The integral control of this interpolation is governed by one cohomological $\mu$-invariant, which depends only on $E[2]$ and vanishes for abelian two-division fields. Together with Kato's cyclotomic divisibility and the fixed lattice in his construction, it bounds the denominator by a fixed power of $2$, independently of the twisting primes and the family dimension. Congruences among coefficients of a half-integral-weight form, in the style of Ono–Skinner, construct the required twist families.
Read the full abstract
Machine Learning
Actions Should EARN Their Credit: Test-Time Adaptation and Route Consolidation for Frozen LLM Agents
LLM agents can use feedback from repeated attempts to improve their decisions without updating model parameters. However, positive action credits can continue to favor actions that repeatedly achieve the same score, even when higher scores are achievable. Under a fixed interaction budget, trying new actions in the final attempts can also lower performance. We introduce EARN, a method that uses past experience to guide further progress and retain earlier performance through conditional reuse of effective action sequences. EARN reduces action credits for reaching scores repeatedly attained in earlier episodes and assigns larger credits to action segments that reach higher scores. These credits adjust the probabilities of LLM-proposed actions. During the final episodes, EARN directly replays stored actions when the current context matches the recorded context and the actions remain available, bypassing LLM candidate generation. We evaluate performance at the end of the interaction budget using the average score over the final four episodes. Across Jericho, ScienceWorld, and Crafter, EARN achieves the best score on this metric in 31 of 33 task–budget settings among the evaluated methods, including ties. Across the evaluated Jericho budgets, replay bypasses 78.9–86.3% of candidate-generation calls in the final four episodes, averaged across games.
Read the full abstract
Sound
From Sound to Occupancy: Privacy-Aware Audio Sensing for Buildings
Occupancy-based controls regulate lighting and HVAC according to whether a space is in use. Passive infrared (PIR) sensors, the most common choice, respond to motion and can report an occupied room as vacant when its occupants remain still. Sound offers an alternative cue, since seated occupants type, turn pages and cough, but microphones raise privacy concerns. We therefore study a two-stage design: an on-device encoder maps each second of audio to a short embedding, and a server-side temporal model turns the stream of embeddings into an occupancy state. To our knowledge, no publicly available dataset pairs office audio with occupancy labels, so we evaluate the two stages separately on public data. The encoder is trained on synthetic scenes that mix empty-room recordings, room impulse responses, office sounds and read speech, and tested on unseen rooms, recordings and speakers for whether a second contains a sound made by a person. On 11 real office recordings (DCASE 2013) it reaches an AUC of 0.87, against 0.76 for log-mel statistics. Attacks on the transmitted embeddings recover at most 1% of the spoken words, against 94% from log-mel frames, but still identify the speaker among 40 in 69 to 77% of the utterances. Noise-floor normalization with adversarial training, an 8-dimensional embedding, or sending only class probabilities lowers this to 38 to 55% without hurting detection. On HPDmobile, which pairs home occupancy labels with an amplitude envelope of the audio, a causal temporal model pretrained with a hierarchical contrastive loss matches hand-crafted features (AUC 0.75) and does better when only one labelled day per home is available. These results make the two-stage design a practical basis for privacy-aware acoustic occupancy sensing, with speaker identity as the one leak still to close.
Read the full abstract
Geophysics
HydroTuring: A physics probe benchmark testing mass, energy, and momentum conservation in hydrologic models
Hydrologic models, including deep learning systems now deployed for operational flood forecasting, are evaluated primarily by how closely their simulated hydrographs match observed streamflow. However, high hydrograph skill does not guarantee physical consistency: in a synthetic calibration experiment, a conceptual model that loses 15% of its water input to an unreported sink still reaches a Nash-Sutcliffe efficiency of 0.93, and no shared benchmark tests whether models obey fundamental conservation laws. Here we introduce HydroTuring (v0.1.0), a black-box benchmark of 37 physics probes that test mass (24 probes), energy (7) and momentum (6) conservation in hydrologic models. Each probe runs a controlled experiment on forcing generated at run time from a recorded seed and applies binary pass or fail criteria validated against 18 physically exact and 59 deliberately broken reference models. Across 32 evaluated models (3 exact references, 7 AI-based models and 22 process-based models), only the 3 exact references pass every probe among the 28 models scored on 5 or more probes. Process-based and AI-based models pass similar shares of their scored probes (76% and 74%), and a pre-trained flood-forecasting LSTM fails 6 of the 11 runoff-only probes it can be scored on. Furthermore, 49% of all 1,184 model-probe pairs remain unscored because models omit the required state or flux variables, while the scored failures occur in perturbation responses and state bounds rather than long-term water-budget closure. HydroTuring provides a reproducible smoke test for model development, peer review and operational deployment under an open community protocol.
Read the full abstract
Machine Learning
UTOPI: Efficient Egocentric Long-Video Understanding in AR via User-Guided Token Pre-Compression
As long-video understanding gains increasing attention, augmented reality (AR) devices are emerging as a compelling deployment platform. However, egocentric videos exhibit significant temporal redundancy, imposing substantial memory and computational burdens on resource-constrained AR hardware. To address this challenge, we introduce UTOPI, a plug-and-play pre-compression module designed specifically for egocentric long videos on AR platforms. First, we leverage user motion signals to estimate inter-frame overlap and propose a viewpoint-shift-invariant pruning algorithm. Second, we exploit eye-tracking sensors to identify user fixations, enabling the separation of attended foreground content from background and its preferential retention via a fine-tuned neural network-based adjustment mechanism. Across multiple benchmarks and model backbones, UTOPI achieves up to 95% compression while preserving performance. Code and project page are available at https://sai-lab-nyu.github.io/utopi/.
Read the full abstract
Human-Computer Interaction
From Sketch to Design: A Cross-scale Workflow for Procedural Generative Urban Design
Urban designers use sketches and symbols to express spatial intentions, while many generative urban design tools rely on numeric parameters that are difficult to relate to this visual process. We propose a cross-scale workflow and a web-based prototype that translates symbolic sketches into controllable procedural urban generation. A vocabulary of shapes, informed by examples of urban design diagrams and sketching exercises with graduate urban design students, encodes attributes such as network structure, density, and program. The prototype uses the \$1 Unistroke Recognizer to map gestures to symbols and translates their locations and sizes into tensor-field attractors and weighted parameter grids. Grid and radial tensor fields guide street-network generation, while the parameter grids control building heights and programs. Implemented in JavaScript and TypeScript with React and Three.js, the prototype supports interactive sketching, visualization, and urban massing generation. We describe the methodology and illustrate prototype interactions, then discuss downstream integration with architectural-scale generative tools for further refinement and analysis. The work connects designers' symbolic input with an explicit procedural generation process.
Read the full abstract
Artificial Intelligence
LiveAsset: Learning Intent Assets from Corrective Interactions for Agent Runtime Control
Large language model (LLM) agents acting across web applications and online services often need multi-turn corrections to align their work with users' underspecified goals and preferences. In practice, users may identify problems in the current outcome without specifying the desired result, so learning user intentions requires interpreting each correction against the agent's preceding and revised outcomes. These expectations are often connected to a particular context, making it hard to know when they apply and how to ensure the agent later acts as the user expects. To this end, we propose LiveAsset, an intent harness that learns reusable intent assets from corrective interactions and uses them for runtime control of agent execution. LiveAsset jointly interprets each correction with the agent's original and revised results to infer user intent, incrementally updating the assets while preserving earlier requirements that remain applicable. The learned requirements guide agent execution from task entry and provide criteria for runtime verification, which uses the explanations to assess applicability before each reply is delivered and directs the agent to revise the result to address unmet requirements while preserving those already satisfied. Following a learning phase in which simulated users progressively reveal requirements through multi-turn corrections, we evaluate LiveAsset using same-task replay of the original requests on {\pi}-Bench, PERMA, and Vita and cross-task transfer on PAST-Bench and SkillEvolBench. Experiments show that LiveAsset better fulfills requirements learned from corrective interactions and improves performance on these benchmarks compared with memory, skill, and rule-based harness baselines. Code is available at github.com/landian60/LiveAsset.