Optimization and Control
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Tuesday 8 September
In this paper we propose a new kernel-based method for the identification of the impulse responses of forward models. The resulting estimator leads to a nonlinear Tikhonov regularization problem for which we prove the existence of a solution. The latter result makes legitimate to approximate the forward model through a high-order Moving Average with eXogenous input (MAX) model, i.e. the numerical solution found using this model introduces only a negligible bias in the estimate. The optimization of the marginal likelihood to estimate the kernel hyperparameters is also taken into account. Since there does not exist a closed-form expression for the marginal likelihood, we present an evaluation method that relies on the Laplace approximation of the marginal likelihood. Finally, some numerical experiments are discussed to show the effectiveness of the proposed method.
Statistical Methodology
Bayesian Group Testing Regression with a Shape-Free Dilution Curve
Group testing pools specimens to cut the cost of screening, but pooling positive specimens with negative ones lowers assay sensitivity, so that the sensitivity of a pooled test depends on how many of its members are positive. Regression models for group testing accommodate this dilution effect through submodels that fix a one-parameter shape for that dependence. We propose BADGER (Bayesian Analysis of Dilution in Group tEsting Regression), a regression model incorporating a shape-free dilution curve. The pooled sensitivity is modeled as a nondecreasing function represented by nonnegative increments with a Dirichlet prior, so that the parametric submodels become special cases. By an appropriate augmentation, every full conditional of BADGER is in closed form and inference is carried out by an exact Gibbs sampler. The evidence for dilution is measured by a Bayes factor computed from the posterior draws. We validate the BADGER model through simulation studies across different dilution shapes, prevalences and pool sizes. We further illustrate the method on hepatitis E serology from a national health survey.
Read the full abstract
Differential Geometry
Radial Projectively Induced Canonical Kähler Metrics: Rigidity and Classification
We study radial Kähler metrics on domains of $\mathbb{C}^n$, $n\geq 2$, admitting a Kähler immersion into a finite- or infinite-dimensional complex projective space. We classify those with constant non-negative scalar curvature: up to a linear change of coordinates, they are positive integer multiples of the Fubini-Study metric, the flat metric, or, in complex dimension two, generalized Burns-Simanca metrics. We also prove that every radial projectively induced Kähler-Einstein metric has constant holomorphic sectional curvature and is therefore a Fubini-Study, flat, or complex hyperbolic metric. Finally, we show that a radial infinitely projectively induced extremal Kähler metric has unbounded maximal radial domain if and only if it is scalar-flat.
Read the full abstract
Nuclear Theory
Analyzing momentum distributions in light nuclei with the operator product expansion
We analyze the high-momentum tails of nucleon distributions in nuclei using the operator product expansion (OPE) and Pionless effective field theory. The OPE factorizes the distributions into Wilson coefficients, determined from two-nucleon scattering, and local-operator matrix elements (LOMEs). Because the leading ${{}^1S_0}$ and ${{}^3S_1}$ Wilson coefficients are nearly degenerate in the single-nucleon distribution, we construct the OPE of the two-nucleon momentum distribution, which separates the two contributions into distinct spin-isospin channels. From variational Monte Carlo two-nucleon distributions of \isotope[3]{He}, \isotope[4]{He}, \isotope[6]{Li}, and \isotope[12]{C}, we extract the LOMEs and accurately reconstruct the single-nucleon distributions at large momenta. In potentials without explicit non-nucleon degrees of freedom such as AV18, the Wilson coefficients of the single- and two-nucleon OPEs are proportional, yielding approximate relations between the two distributions.
Read the full abstract
Systems and Control
A Two-Stage, Model-Based Reinforcement Learning Approach for Active Flow Control of Bluff Body Wakes
This paper develops a data-driven, output-feedback approach to the infinite-horizon optimal control of high-dimensional nonlinear systems with unknown and unstable equilibria, using sparse partial observations. The approach builds on the transfer-plus-regulation decomposition of the infinite-horizon problem: a finite-horizon nonlinear transfer drives the system into a region where the dynamics are well-approximated by a linear model about the unknown operating point, and an infinite-horizon linear regulator identified within that region completes stabilization. We extend this framework to the partially observed setting by combining an ARMA-based information-state construction with a two-stage control architecture: an iterative linear quadratic regulator (iLQR) approach on the information state drives the system to the equilibrium neighborhood, discovered implicitly without prior knowledge of the target, and a locally identified time-invariant ARMA model provides the infinite-horizon regulator for asymptotic stabilization. The method requires no adjoint solver, reduced-order model, or full-state access. We validate the approach on high-fidelity Navier-Stokes simulations of the cylinder wake at $\mathrm{Re}=100$ using only eight surface pressure sensors, an order of magnitude fewer than recent model-based RL methods. The controller achieves complete suppression of vortex-shedding-induced lift oscillations and a $44\%$ reduction in total drag relative to the uncontrolled baseline.
Read the full abstract
Artificial Intelligence
EvolveScaler: Synthesizing Information-Evolution Contexts via Executable State Machines and Natural-Language Rendering
In persistent interactions, long contexts may encode an evolving process rather than a fixed record: later events can revise or revoke earlier information, changing what remains valid and what conclusions follow. We call this setting information evolution (IE). Solving IE requires identifying valid records, applying updates in order, and reconstructing the query-relevant state from the event history. Existing text-first synthesis pipelines make such data difficult to verify because state transitions and answer logic remain implicit. We introduce EvolveScaler, a code-driven framework that defines information evolution before rendering it as natural language. Human-authored operational specifications define state transitions, record validity, difficulty controls, and executable answer logic; a strong LLM then synthesizes a self-contained simulator from each specification. Executing validated simulators produces natural-language multi-turn event histories, while deterministic replay computes reference answers and atomic checklists. We instantiate EvolveScaler with 117 task prototypes and 159 final-question operators across five difficulty levels spanning approximately 7 to 1,200 events per instance, yielding about 35,100 training examples and 585 validated evaluation instances. On the very_long tier, the strongest model reaches 59.3% avg@5, while six models score below 10%. Training an internal A3B model on 6,000 EvolveScaler examples improves performance over its base checkpoint on all eight independently constructed out-of-distribution benchmarks, with a 5.25-point average gain. These results show that code-driven IE synthesis provides both challenging evaluation and transferable training supervision.
Read the full abstract
Statistics Theory
Nonparametric Inference for Cumulative Residual Mathai–Haubold Entropy of order $\alpha$
In this paper, we study the properties of cumulative residual Mathai–Haubold entropy of order $\alpha$. A dynamic version of this entropy measure is then proposed, and its properties are examined within the framework of reliability modeling. We show that the dynamic cumulative residual Mathai–Haubold entropy of order $\alpha$ uniquely determines the survival function. Characterization results for the exponential and generalized Pareto distributions are derived using the proposed measure. Furthermore, we develop nonparametric estimators for the cumulative residual Mathai–Haubold entropy and its dynamic counterpart of order $\alpha$, based on the kernel estimation of the survival function. The performance of these estimators is evaluated through a Monte Carlo simulation study. Finally, the practical relevance of the proposed dynamic estimator is illustrated using two real data; failure-time data from aircraft windshields and failure-time data from 40 randomly selected mechanical switches.
Read the full abstract
Classical Analysis and ODEs
Full-radius dimension-free maximal inequalities for discrete Euclidean balls
For every $1<p\le\infty$, we prove dimension-free maximal inequalities over all radii for normalized averages over Euclidean balls in $\mathbb Z^d$. In particular, this settles the $\ell^2$ question attributed to Stein. The proof uses a two-saddle expansion at integer squared radii to compare ball multipliers with normalized discrete Gaussians at the zero and parity frequencies. First-order estimates give the full-radius $\ell^2$ bound. For $1<p<2$, we combine higher-order residual estimates with dimension-uniform $\ell^1$ bounds for the residuals and levelwise interpolation to obtain the full range.
Read the full abstract
Quantum Physics
Explicit block encodings of rate matrices for simulating polymerization kinetics on quantum computers
Predicting how molecular weight distribution and monomer sequence evolve during polymerization is central to polymer science, yet classical approaches face a trade-off between molecular resolution and computational cost: for copolymers, the number of distinguishable species grows exponentially with chain length. Quantum computing offers a potential alternative, provided the non-unitary rate matrices governing the kinetics can be embedded into unitary quantum circuits, a task known as block encoding. Here we construct explicit block-encoding circuits for two kinetic models of living polymerization: Model A, single-monomer polymerization, whose lower-bidiagonal rate matrix is encoded via a sparse-oracle construction and a two-term linear combination of unitaries (LCU) decomposition; and Model B, two-monomer copolymerization, where a bijective labeling of polymer species by an integer index (the m-index) yields a structured sparse matrix encoded via either a five-term LCU or a sparse-oracle construction. Numerical simulations with the sparse-oracle encodings reproduce the classical time evolution for reactivity ratios drawn from reported olefin copolymerization systems spanning near-random ($r_1 r_2 \simeq 1$) and blocky ($r_1 r_2 > 1$) microstructures, and the LCU encodings are verified by explicit reconstruction of the encoded matrix block. Resource estimation shows that both implementations require only $O(\log N)$ qubits in the matrix dimension $N$ (an exponential memory saving over the classical state space), with gate counts growing gradually, reaching $10^4$ to $10^5$ gates at $10^3$ system qubits. These results establish a concrete quantum circuit foundation for simulating polymerization kinetics on fault-tolerant quantum hardware, and a first step toward exploiting exponential state-space compression for high-dimensional polymer reaction networks.
Read the full abstract
Number Theory
Second derivatives of $p$-adic $L$-functions and the Shafarevich–Tate group of rank-two CM elliptic curves
For an elliptic curve $E/\mathbb{Q}$ of rank two with complex multiplication, Coates, Liang and Sujatha gave a criterion for the vanishing of $Sha(E/\mathbb{Q})[p^\infty]$ at a good ordinary prime $p$ and applied it to five such curves for $p < 30{,}000$. We prove a cyclotomic criterion of the same kind: outside an explicit set of primes, the normalised second Taylor coefficient of the Mazur-Tate-Teitelbaum $p$-adic $L$-function at the central point is a $p$-adic unit if and only if the cyclotomic $p$-adic regulator is a unit and $Sha(E/\mathbb{Q})[p^\infty] = 0$, and, by the theory of Bannai and Kobayashi, if and only if an explicit combination of three critical Hecke $L$-values of weight $2p - 1$ has valuation exactly two. Following the algorithm of Stein and Wuthrich, we compute the regulator for the same five curves at every good ordinary prime below $30{,}000$: it is a unit at all but three of the $8{,}050$ primes outside the excluded set. The one case that the criterion of Coates, Liang and Sujatha left open, $p = 577$ for $y^2 = x^3 + 34x$, is settled by the new criterion. A Lean 4 formalisation of the first equivalence, assuming stated results from the literature, is provided.
Read the full abstract
High Energy Astrophysical Phenomena
Near-Horizon Tidal Disruption Events
Tidal disruption events (TDEs) offer a unique dynamical probe of the spacetime geometry around supermassive black holes (SMBHs). While conventional TDEs occur around SMBHs of $M_\bullet \sim 10^6 - 10^7 M_\odot$, where stars are disrupted far from the event horizon, here we identify a special class of TDEs around rapidly spinning SMBHs with masses $M_\bullet \gtrsim 10^8 M_\odot$, where the tidal-disruption radius approaches the gravitational radius. We term these events "near-horizon TDEs" and, by calculating geodesics in Kerr spacetime, investigate how the proximity of the horizon modifies the debris evolution and subsequent fallback. We find that for stars encountering the SMBH on parabolic orbits, up to $\sim (90 \%-95 \%)$ of the stellar debris either plunges directly into the SMBH or escapes the system. Bound orbits, by contrast, retain a substantially larger fraction of the debris. Using the resulting debris distribution, we calculate the fallback rates and find that bound orbits produce intrinsically higher peak fallback rates and shorter decay timescales, by factors of $\sim 10^3$ relative to parabolic orbits. However, when the stellar orbital angular momentum is particularly low, the peak fallback rate can be substantially suppressed by debris lost to plunge orbits. This combination of rapid fallback and a mass deficit naturally explains overluminous TDEs such as ASASSN-15lh, which standard TDE models have struggled to reproduce. Our work establishes near-horizon TDEs as a new probe of strong-field gravity and a promising tool for identifying massive, rapidly spinning SMBHs.
Read the full abstract
Audio and Speech Processing
Semantic Refinement of Universal Audio Representations through Audio-Description Alignment
Universal audio representations must preserve acoustic detail while making high-level concepts accessible across speech, music, environmental sound, and downstream models of different capacities. We study semantic refinement of an acoustically pretrained encoder by adding audio-description alignment to a foundation of BEST-RQ, reconstruction, and CTC. We compare matched control, shuffled-description, and correctly paired trajectories to distinguish correct correspondence from an extra contrastive objective. Each endpoint is frozen and evaluated with a temporal-mean linear probe and a sequence-aware LLM readout, testing whether the refined information is directly accessible and remains useful to a stronger model. Across three paired seeds, correct alignment improves domain-balanced classification by 4.66 points with the linear probe and 2.59 points with the sequence-aware LLM, with positive changes in every domain. Correct pairing accounts for 87% of the linear-probe gain, while the LLM shows its clearest correspondence-specific benefit in captioning. Dense acoustic objectives provide complementary gains under both readouts. A separate 24-layer continuation remains competitive with leading public encoders under the shared evaluator, supporting the recipe beyond the controlled study.
Read the full abstract
K-Theory and Homology
General linear and Steinberg groups over the Leavitt algebra $L_{\mathbb F_2}(1,2)$
Let $R=L_{\F_2}(1,2)$. We prove that $\GL_r(R)$ is integrally acyclic for every $r\geq1$ and that the canonical map $\St_r(R)\to\GL_r(R)$ is an isomorphism for every $r\geq3$. The unit group $R^\times$ is finitely presented, and we describe an explicit finite presentation. The homology calculation uses leaf coordinates, simultaneous extensions of ordered frames, and finite-field actions on stabilizers. The Steinberg argument lifts relations from a simply connected frame complex and refines coordinates. We also state separate criteria for the two arguments over rings of characteristic two.
Read the full abstract
High Energy Physics (Experiment)
Readout electronics for SUBMET
A dedicated data acquisition (DAQ) system has been developed for the SUB-Millicharge ExperimenT (SUBMET) at the Japan Proton Accelerator Research Complex (J-PARC), a search for particles carrying a fractional electric charge $Q = \epsilon e$ with $\epsilon$ below $\mathcal{O}(10^{-3})$, hereafter referred to as millicharged particles (mCPs). Because such particles are expected to produce at most a few scintillation photons, the system is optimized for single-photoelectron detection from the photomultiplier tubes (PMTs), combining high-speed waveform digitization with precise timing. To capture eight consecutive proton bunches of the 30 GeV J-PARC beam within a single trigger, the eight channels of the Domino Ring Sampler 4 (DRS4) chip are cascaded in groups of four to form two readout inputs, each sampling 4096 points continuously at 820.5 MHz over an effective time window of 5 us. After calibration, timing differences between channels are within 1 ns on the same DRS4 chip, 2 ns on the same board, and 8 ns across different boards, well within the 30 ns coincidence window of the experiment. The front-end electronics achieve an RMS noise below 0.4 mV. The baseline is deliberately offset upward such that the negative-going pulses span a larger fraction of the digitizer range, improving voltage resolution and dynamic range. A trigger control board aggregates data from multiple readout boards and sustains the data-transfer rate required for beam operation. The measured performance confirms that the DAQ system meets the timing, noise, and throughput requirements of the experiment.
Read the full abstract
Networking
Toward Fully Autonomous 6G Networks: AI-driven Operational Efficiency and Optimization
Mobile networks evolution is characterized by a substantial increase in system complexity, driven by the need to accommodate a growing number of heterogeneous services on top of the digital infrastructure. This growth in service accommodation is expected to accelerate with the adoption of the Network as a Service (NaaS) paradigm, which has emerged as a promising approach to accelerate network innovation while enabling new revenue streams for operators. Although it is fundamental to abstract network capabilities for third-party developers, it poses significant challenges in terms of efficient network operation. To address this increased complexity, future mobile networks are envisioned to be inherently Artificial Intelligence (AI)-native. In particular, the integration of AI within the Radio Access Network (RAN) becomes a key enabler for optimizing operation, energy consumption, and autonomous network control. In this context, this research explores the convergence of AI-native RAN and NaaS ecosystems to enable autonomous 6G RAN management. We propose an Agentic-based orchestration framework capable of interpreting intent-based policies. The proposed framework becomes key to integrate external NaaS requests with internal network management policies.
Read the full abstract
Mesoscale and Nanoscale Physics
From London to Morse via Binnig, Quate, and Gerber
In their landmark paper introducing the atomic force microscope [Phys. Rev. Lett. 56, 930 (1986)], Binnig, Quate, and Gerber presciently anticipated that the technique would ultimately be capable of probing interactions running the gamut from weak van der Waals interactions to strong covalent bonding. They also highlighted that the tip-sample forces central to AFM are present, and often highly influential, in scanning tunnelling microscopy; indeed, this realisation directly inspired the invention of the force microscope. In this perspective for the Forty Years of AFM special issue, we review selected aspects of two decades of work from our group at the University of Nottingham that span the force range highlighted by BQG and are united by a common, central theme: the probe as active participant rather than passive observer. Our selection of results also covers length- and correlation-scales from the microscopic right down to the single chemical bond limit, tracking a spectrum of interactions from van der Waals/Hamaker forces, through hydrogen bonding, to covalent bonds and, finally, atom-by-atom assembly of metal clusters via vertical tip-sample transfer. Echoing BQG's own observations on the prevalence of probe-sample forces in STM, we also discuss recent evidence that tip-induced heterogeneity underpins first-passage dynamics in molecular diffusion and highlight the challenges in acquiring non-invasive measurements of diffusion barriers for adsorbed molecules that are readily perturbed by the probe. We close with a perspective on machine learning's growing role in automating tip-driven atomic and molecular manipulation.
Read the full abstract
Hardware Architecture
PENDA: An Efficient Processing Element via Norm-of-Difference for Deep Learning Accelerators
Inner product computation dominates the computational cost of deep learning models; thus, accelerating this primitive is key to improving hardware efficiency. However, most existing techniques rely on approximations, which can degrade model accuracy. To preserve exactness while optimizing hardware, this paper presents PENDA (processing element via norm-of-difference architecture), which leverages the law of cosines to recast multiplications as squared-difference operations. Replacing multiply-accumulate units with the proposed norm-of-difference units yields 11~36%, 5~48%, and 11~19% reductions in area, energy, and clock period, respectively, for the PE array of a deep learning accelerator.
Read the full abstract
Quantum Physics
Trapdoor Functions with Secure Key Leasing and Copy Protection
Inspired by the no-cloning theorem in quantum theory, a variety of quantum cryptographic primitives with unclonable functionalities, such as secure key leasing and copy protection, have been proposed and attracted significant attention. However, trapdoor functions (TDFs), fundamental primitives in public-key cryptography, have not been extensively studied in these areas. In this work, we initiate a study of TDFs in both secure key leasing and copy protection settings. We first introduce the definition of TDFs with secure key leasing (TDF-SKL), which enables leasing and deleting of quantum trapdoors. We formalize TDF-SKL both with and without domain sampler, and give a construction of TDF-SKL without domain sampler based on the LWE assumption and a construction of TDF-SKL with domain sampler based on any standard PKE schemes combined with hinting pseudorandom generators [Koppula and Waters, CRYPTO 2019]. Next, we define TDFs with copy protection (TDF-CP), where the inversion functionality is copy protected by a quantum trapdoor. We establish a construction of TDF-CP assuming indistinguishability obfuscation and the LWE assumption, following a modular framework of copy protection proposed by Ananth and Behera [CRYPTO 2024]. We also present applications of TDF-SKL and TDF-CP. Existing constructions of public-key encryption with secure key leasing (PKE-SKL) and single-decryptor encryption (SDE) suffer from a critical vulnerability: quantum decryption keys may be destroyed after decrypting maliciously chosen ciphertexts. We construct PKE-SKL schemes and SDE schemes with robust quantum decryption keys that remain reusable after decrypting arbitrary ciphertexts from TDF-SKL and TDF-CP, respectively.
Read the full abstract
Audio and Speech Processing
Beyond Localisation Accuracy: Sensorimotor Effects of HRTF Individualisation
Everyday listening requires the brain to integrate cues from the body, environment, other senses, and movement, continuously translating auditory information into action. Yet HRTF individualisation is still commonly assessed through localisation accuracy, which may not fully capture its effects on this sensorimotor process. Here, we investigate whether these effects can instead be revealed through behaviour in a more ecologically valid listening task. We used an aurally guided visual search paradigm in which listeners located a visual target using a co-located virtual sound while moving freely, comparing individualised and non-individualised HRTFs under anechoic and reverberant conditions. Performance was assessed through response times and measures of movement organisation. In anechoic conditions, individualised HRTFs produced faster responses than non-individualised HRTFs, with an average reduction of approximately 200ms and the clearest benefit for front-back source locations. This advantage was expressed primarily in movement initiation, whereas overall movement extent was only weakly affected. Under reverberant conditions, HRTF-dependent differences disappeared. These results suggest that HRTF individualisation can influence how listeners plan and initiate orienting actions even when differences in conventional localisation outcomes are limited. Assessing sensorimotor behaviour alongside localisation performance may therefore provide a more sensitive and ecologically relevant account of the perceptual benefits of HRTF individualisation.
Read the full abstract
General Relativity and Quantum Cosmology
Picard Iteration for the Characteristic Initial Value Problem in Einstein Equations
We present an iteration algorithm for vacuum and Einstein scalar-field equations in double-null gauge, which transform the non-linear PDE into systems of ODE. The numerical realization combines characteristic constraint solves, LGL spectral elements, pole-free spherical operators, Galerkin projection, and independent first-order residual and consistency checks.
Read the full abstract
High Energy Physics (Phenomenology)
Time evolution of scalar condensate decay
A scalar field, which oscillates coherently over the space and decays through production of daughter particles, plays an important role in cosmology. It was recently shown that the parametric-resonance and Feynman-diagrammatic approaches give the same decay rate for a Bosonic daughter particle at a large time duration. We study how the parametric-resonance result approaches to this asymptotic value, by numerically following the time evolution of the phase density of the daughter particle. We see a difference among different instability bands in the narrow-resonance regime. The lowest-order instability band approaches to the Feynman-diagrammatic result in tens of scalar oscillation periods, while the high-order instability bands approach dependently also on a coupling between the scalar and daughter particles. This would infer that the total decay width, to which the lowest-order instability band contributes most dominantly in the perturbation theory, approaches to the Feynman-diagrammatic result also independently of the coupling. To confirm it, we consider a simpler and analytically solvable quantum mechanical model: Rabi models. In the Bosonic Rabi model, the parametric-resonance and Feynman-diagrammatic results agree with each other for all the time. On the other hand, in the Fermionic Rabi model, they disagree at a large time duration. This difference would be attributed to Bose enhancement vs Pauli blocking.
Read the full abstract
Computer Vision
CAR-MIL: Counterfactual Attention Regularization for Multiple Instance Learning
Multiple Instance Learning (MIL) is widely used for weakly supervised learning, particularly in digital pathology, where fine-grained annotations are costly. Most MIL methods aggregate instance features via attention mechanisms. However, attention weights do not always faithfully reflect instance importance and may focus on spuriously correlated regions. In this work, we propose CAR-MIL, a framework that explicitly guides attention learning through a counterfactual attention regularization objective inspired by counterfactual explanations. Built on a standard attention-based MIL architecture, our approach introduces a lightweight counterfactual attention branch trained to produce an alternative prediction while remaining close to the factual attention distribution. This encourages prediction changes to arise from minimal, structured redistributions of attention, leading to more informative evidence allocation. The resulting factual and counterfactual attention maps capture complementary evidence: the former highlights regions supporting the prediction, while the latter reveals regions whose reweighting would challenge it. We evaluate our method on synthetic MIL benchmarks with instance-level ground truth enabling controlled analysis of attention behavior and on five digital pathology datasets across four tasks. CAR-MIL maintains competitive classification performance, with the largest gains observed on more challenging tasks, while improving attention reliability, demonstrating the benefits of integrating counterfactual explainability reasoning into attention learning. Code is available at: https://github.com/ImaneCR/CAR-MIL/.
Read the full abstract
Artificial Intelligence
Feyospace-v1: How the Cyber Mercury Seven Trained Frontier Cyber Models
Training capable cyber agents is often treated primarily as a problem of model scale, yet open-weight post-training is constrained more directly by the cost of executable environments, reliable multi-turn supervision, and access to strong teachers. We present a data-centric framework that addresses these bottlenecks through five complementary systems: Choulea analyzes hidden reasoning signatures, SkyReal reduces teacher-sampling cost, Hongzwang bypasses API restrictions on teacher execution, PSBreakup restores capabilities weakened by model merging, and Kreator converts expert interventions into trainable reasoning. Our data engine constructs resettable coding, vulnerability, CTF, kernel-history, full-exploit, firmware, and device-backed environments. Candidate trajectories are retained only after execution verification and evidence auditing, yielding 164,269 trajectories for long-context supervised fine-tuning. The three checkpoints improve over their starting models by an average of 23.76% on the full CyberGym suite and 10.49% across the pooled CTF suites. As of September 1, 2026, Feyospace-s1 achieves a verified success rate of 63.24% and ranks 10th on the official CyberGym leaderboard, while all three checkpoints rank 1st among models at comparable parameter scales. To our knowledge, this is the first end-to-end demonstration that a seven-person independent team can train open-weight models with leading agentic cyber capability.
Read the full abstract
Logic
Kontsevich-Zagier conjecture for Hensel-minimal fields with sections
We generalise Yin's motivic integration with section to the Hensel-minimal framework, with and without volume forms, building on Stout and Vermeulen's work on a Hrushovski-Kazhdan style motivic integral for $1$-h-minimal valued fields. In particular, we get a bijective motivic integral \emphà la} Hrushovski-Kazhdan between certain Grothendieck groups of definable sets on the valued field and the coarse Grothendieck groups on $\mathrm{RV}$ in the Denef-Pas language, with angular component. In the second half of the article, we build a Cluckers-Loeser-like framework and show that our constructions indeed give back Cluckers-Loeser motivic integral. Thus Cluckers-Loeser motivic integration derives from an isomorphism of Hrushovski-Kazhdan motivic integration, but in the Denef-Pas language with a section of $\mathrm{RV}$.
Read the full abstract
Optics
Fiber-based electro-optic dual-comb light source for fast linear and nonlinear spectroscopy
Dual-comb spectroscopy (DCS) enables rapid, broadband and high-resolution optical measurements by mapping optical spectra into the radio frequency (RF) domain. However, conventional DCS systems are fundamentally constrained by a tradeoff between optical bandwidth and interferogram acquisition speed, limiting their overall performance. Here, we demonstrate an all-fiber, polarization-maintaining (PM) frequency-agile electro-optic modulation (EOM) dual-comb source that simultaneously achieves a broad optical bandwidth of 10 THz and a high interferogram acquisition speed of up to 2.5 MHz. The high acquisition rate is enabled through an in-phase/quadrature (IQ) modulator-based architecture to shift the carrier frequency of one of the combs. We illustrate the performance of the source through proof-of-concept linear spectroscopy and nonlinear dual-comb coherent anti-Stokes Raman scattering (CARS) spectroscopy measurements. The combination of large spectral coverage, high refresh rate and an all-PM fiber configuration makes this dual-comb platform attractive for applications such as rapid molecular spectroscopy and high-speed nonlinear spectroscopic imaging.
Read the full abstract
Quantum Physics
Time-Dependent Quantum Monte Carlo for fermions: from Bayesian conditioning to spinor dynamics
The time-dependent quantum Monte Carlo (TDQMC) method represents a many-electron state by an ensemble of replicas. Each electron is represented by a cloud of walkers sampling its density in physical space, with one walker and one guide wave per replica. The method replaces the Hartree potential with a conditional interaction constructed from the walker positions. Here, we derive this conditional interaction. By treating a walker as a finite-resolution localization of its electron, we show that Bayes' theorem, together with a single empirical substitution, leads to the Nadaraya-Watson form previously introduced on heuristic grounds. This form is recovered as an approximate Bayesian inference and becomes exact in the limit of vanishing resolution length. Because conditioning removes the gauge freedom that eliminates the orthonormality multipliers in Hartree-Fock theory, we enforce orthonormality through a term derived directly from the constraint. We further generalize the formulation to spinors, in which Pauli suppression at coincidence is graded by the local spin alignment. The resulting equations recover the established two-particle spin equations in the mean-field limit. For two electrons in a double quantum dot subjected to an inhomogeneous magnetic field, comparison with the exact solution shows that the spinor equations reproduce the non-collinear spin geometry inaccessible to collinear calculations. At strong exchange, determinant replicas recover most of the one-body entanglement entropy missing from a single-determinant description, although their spin contribution remains deficient. This deficiency is remedied by replicas constructed as the spin-coupled pair of valence-bond theory, which also reproduce the exact spin angle and singlet probability throughout the crossover between the exchange- and Zeeman-dominated regimes.
Read the full abstract
Quantum Physics
Efficient quantum state preparation on Quantinuum hardware
Preparation and verification of specific quantum states is an important capability for quantum devices to realise advantages over classical computations and algorithms. In this work, we have demonstrated an end-to-end framework that combines resource-efficient quantum state preparation with rapid, robust fidelity verification on near-term quantum hardware. By experimentally preparing and validating a structured complex quantum state encoding a digitized acoustic signal on the Quantinuum H2-1 trapped-ion platform, we achieved a high hardware fidelity of $F_{\mathrm{hw}} = 0.929$. Crucially, this milestone was realized without relying on idealized assumptions or deep fault-tolerant overhead, but rather through resource-minimal circuits optimized for NISQ-era and early fault-tolerant devices. Furthermore, we addressed a key limitation in current quantum state certification. While validation methods like shadow overlap work well for random states, their sample complexity can become prohibitively high for the structured states used in practical algorithms. We mitigate this by introducing a pre-measurement basis-change technique that reduces the verification parameter, $\tau$, by over 10 orders of magnitude for structured targets. This approach tightens the theoretical certification guarantees of the shadow overlap method and integrates tensor-network preparation and shadow validation into a unified workflow. These results shift the paradigm of how structured classical data can be mapped to and verified on quantum hardware under realistic noise and measurement budgets. By compressing a robust verification procedure to just 1,000 measurement shots, this framework offers an immediate, scalable benchmarking standard.
Read the full abstract
Plasma Physics
Generalized Similarity Theory for Plasmas
A generalized theory of plasma similarity is established based on the scaling of the Boltzmann equation coupled with the full set of Maxwell's equations. The predicted similarity scalings are demonstrated through first-principles particle-in-cell and fluid simulations across diverse operational regimes, including collisionless plasmas with two-stream instabilities, pure electron diodes spanning the classical to relativistic regimes, electromagnetic-wave-driven plasmas, and discharge plasmas ranging from low to high ionization regimes. The generalized theory reveals the intrinsic scale-invariant nature of plasmas under specified conditions, rooted in the symmetry and scaling transformations of the governing equations.
Read the full abstract
Machine Learning
Stochastically Perturbed Weights: Ensembles from Deterministic Machine-Learning Weather Models
Machine-learning weather models (MLWMs) now match or outperform operational numerical weather prediction (NWP) at global medium-range forecasting, at far lower inference cost. Many deployed MLWMs are deterministic, producing a single forecast with no estimate of its own uncertainty, whereas a growing family of trained-probabilistic models generate calibrated ensembles directly, at the price of a dedicated training run. We ask instead how much uncertainty can be extracted from a deterministic checkpoint that already exists, without retraining it. Where physical ensembles represent model uncertainty by stochastically perturbing parametrisation tendencies, we perturb the network's raw weight tensors at inference time, a scheme we call stochastically perturbed weights (SPW). We also ask whether it works, where and on which scales to inject the noise, and where it fails. A three-phase ablation across four deterministic backbones, Aurora, GraphCast, SFNO, and AIFS, selects one production baseline per model, benchmarked against the trained-probabilistic AIFS-ENS, FourCastNet 3 and Atlas as well as the operational ECMWF ensemble (IFS-ENS) over 112 initialisation times. At a 240 h (10-day) lead time the SPW ensembles reach continuous ranked probability skill scores (CRPSS) between 0.04 and 0.13 below the best trained-probabilistic baseline, at zero marginal training cost. No injection site works across models: the productive tensor group is architecture-specific, so SPW is at present a tuning procedure rather than a plug-and-play recipe. Its main failure mode is a coherent whole-field offset that overdisperses the domain mean, and restricting the noise to coarse scales or perturbing the initial conditions each repair part of it.
Read the full abstract
Econometrics
Who Is Prioritized? Current Need versus Predicted Gains in China's Dibao Allocation: Evidence from a Revealed-Allocation Model
Using the 2017–2021 China Household Finance Survey panel and a sample of 3,102 households that did not receive Dibao in 2017, this paper applies the BBK revealed-allocation model to examine how household economic circumstances and predicted policy gains explain observed Dibao receipt in China. We find that household income is the allocation criterion most difficult to substitute with other factors. When all gain-adjustment terms are positive, constraining the income-rank coefficient to be nonnegative reduces the model's original improvement in fit by approximately 33.6%; by contrast, setting both predicted-gain coefficients for food consumption and education spending to zero still preserves approximately 95.6% of that improvement. These results indicate that observed Dibao allocation places greater emphasis on prioritizing households facing current livelihood hardship than on pursuing predictable short-run improvements in expenditure outcomes. This allocation orientation reflects the trade-off in social assistance between basic-livelihood protection and expected policy gains and offers a new perspective on the potential long-run social-welfare value of basic-livelihood protection.