Debates over open source and openness in artificial intelligence have intensified as policymakers, researchers, and practitioners grapple with how foundation models should be developed and governed to balance innovation, accountability, and public interest. However, there has been limited empirical work examining how diverse stakeholders collectively understand and negotiate responsible openness in AI, particularly through participatory processes that extend beyond industry-led definitions and frameworks. This paper presents findings from a multi-sectoral workshop grounded in futures thinking and participatory design methods. The workshop generated co-created visions of desirable futures and the role of AI, alongside a set of action pathways and a research roadmap focused on responsible open source and openness in AI. This paper makes three key contributions. First, it empirically documents the co-created visions, actions, and research priorities. Second, it identifies four core tensions that emerged as participants translated high-level aspirations into concrete actions, revealing conflicting interpretations of openness regarding its purpose (as an end or a means), its scope (expansion versus meaningful access), and its operation (mandatory versus conditional, sufficient versus dependent on governance and use). These tensions illustrate that responsible openness is not a singular technical solution, but a negotiated sociotechnical project shaped by values, positionalities, and priorities. Third, the paper advances methodological approaches in AI governance by demonstrating how participatory futures methods can surface plural visions, actions, and research priorities that extend beyond dominant, largely corporate, narratives, offering empirical insight into how openness, power, and accountability are negotiated in practice.

The Majda-McLaughlin-Tabak (MMT) family of models has proven to be an efficient ground for benchmarking wave turbulence theory, thanks to the low computational cost required to test theoretical ideas and the possibility of tuning nonlinearity and dispersive properties of the equations. Here, we study numerically the wave kinetic equation (WKE) associated with the MMT model and perform simulations to study turbulent cascades. We confirm numerically the predictions of wave turbulence theory, both in the parameter space region where the wave kinetic equation was proven to be well posed and outside of it. We also observe a new stable stationary state in a region where no cascade solutions are expected, a region that, to the best of our knowledge, has not been explored before. Moreover, following recent work, we study next-to-leading-order corrections to the wave kinetic equation; we uncover incurable divergences in the one-dimensional MMT model and, more generally, in higher-dimensional systems with concave power-law dispersion relations.

DRAM suffers from read disturbance phenomena (e.g., RowHammer and RowPress), where repeatedly accessing or continuously keeping open a DRAM row (aggressor row) induces bitflips in other physically nearby unaccessed rows (victim rows). The disturbance mechanism is practically exploitable from the software stack and worsens across generations with continued density scaling. DRAM read disturbance is highly sensitive to memory access patterns, yet prior work explores read disturbance under only a limited set of access patterns. We present ScaleDisturb, a new DRAM access pattern that can amplify DRAM read disturbance by asymmetrically extending the open time of two aggressor rows. Our rigorous experimental characterization of 196 DDR4 and 3 HBM2 DRAM chips shows that ScaleDisturb (1) leads to bitflips at significantly fewer row activations, compared to state-of-the-art memory access patterns, (2) makes read disturbance attacks easier across all tested DRAM chips, (3) increases DRAM vulnerability to read disturbance as DRAM manufacturing technology scales down to smaller node sizes. We showcase a proof-of-concept attack on a real system where a user-level program leveraging ScaleDisturb induces more bitflips than state-of-the-art RowHammer and RowPress memory access patterns. We describe and evaluate four solutions for mitigating read disturbance bitflips in the presence of ScaleDisturb and call for more research on the topic.

Next-generation Adaptive Optics (AO) systems for 8-40m class telescopes, such as MORFEO (ELT) and MAVIS (VLT), demand high calibration accuracy. Controlling thousands of actuators makes experimental calibration unfeasible due to daytime overheads, environmental disturbances, and opto-mechanical aberrations. Consequently, model-based (synthetic) calibration has become the mandatory standard. We present SynIM, an open-source Python library designed for computing high-fidelity synthetic Interaction, Projection, and Covariance Matrices. SynIM leverages GPU acceleration via CuPy to handle the massive dimensionality of high-order systems. A core innovation is its handling of spatial geometry via composite affine transformations and absolute sub-pixel grid alignment. By merging DM and WFS shifts, rotations, and magnifications into a single operation, SynIM minimizes interpolation artifacts. SynIM introduces an optimized numerical derivative engine for slope computation that mathematically aligns spatial grids at the sub-pixel level, closely mimicking the physical behavior of Shack-Hartmann sensors. It outperforms geometric estimators like the G-tilt telescoping sum at high spatial frequencies, while yielding a substantial computational speed-up. Crucially, end-to-end MCAO simulations demonstrate that reconstructors built with SynIM deliver closed-loop AO performance practically equivalent to full physical optics models. SynIM natively supports SCAO, GLAO, MCAO, and LTAO configurations. It features optimized multi-WFS batch processing, modules for MMSE tomographic reconstructors, and full compatibility with SPRINT for online tracking. Currently driving the design and operational strategies for MORFEO, MAVIS, AOF, KAPA, and WST, SynIM stands as an essential tool for next-generation AO calibration.

Data-driven control of nonlinear systems from input-output measurements remains a fundamental challenge, as existing approaches with rigorous closed-loop guarantees predominantly require access to full state measurements. In this paper, we address this gap by proposing a data-driven output-feedback controller design method for nonlinear systems that provides provable closed-loop guarantees while operating solely on measured input-output data. Our approach combines Koopman operator theory with an extended state representation of the nonlinear system constructed from input-output trajectories. This allows us to obtain a bilinear surrogate model directly from data, on which robust state-feedback design methods can be applied. By exploiting the observability of the underlying nonlinear system, we establish exponential stability of the extended state, which in turn implies exponential convergence of the original system state to the origin. Finally, we validate our theoretical findings in numerical simulations.

We study the Hamiltonian $H_n(h,σ)=\sum_{i=1}^n h_i(σ_i-m), $ where $(h_i)$ are i.i.d.\ real random variables and $(σ_i)$ are i.i.d.\ Ising spins. We consider the energy levels obtained after an independent thinning that retains an exponential number of configurations ($e^{O(n)}$). We prove that, after an $(h_i)$-dependent centering, the resulting point process converges in distribution to a Poisson point process with exponential intensity. Thus, the energy levels asymptotically has the one of the Random Energy Model (REM). Our results extend previous ones, where REM universality for this model was established only either for energy fluctuations of order $e^{-O(n)}$ or for $e^{o(\sqrt n)}$ randomly selected configurations. We also identify the limiting Gibbs weights, which converge to a Poisson--Dirichlet law, and the quenched free energy, which exhibits a freezing transition at $β=\tildeλ$. The proofs are presented here in compressed form; full details are given in the companion preprint.

MORFEO (Multi-conjugate adaptive Optics Relay For ELT Observations, formerly MAORY) is the multi-conjugate adaptive optics module of the ESO Extremely Large Telescope (ELT), designed to deliver diffraction-limited performance in the near-infrared for its first-light camera MICADO. With its 12 wavefront sensors and three deformable mirrors, MORFEO stands as the largest and most complex adaptive optics system ever developed for astronomical observations. A key aspect of its design and verification is the construction of a comprehensive wavefront error (WFE) budget, which defines the system's expected performance under a range of operating conditions. In this work, we present the structure of the MORFEO WFE budget, detailing the main contributors to the residual error and the methodology adopted to evaluate each term. The analysis includes contributions from atmospheric turbulence, optical surfaces, alignment tolerances, control residuals, and calibration uncertainties. We also discuss the external conditions, system configurations, and key assumptions used in the derivation of the budget, highlighting the interplay between requirements and instrument design choices. Importantly, the resulting error budget is being used to support end-to-end simulations of MORFEO-assisted observations, providing essential input for assessing the scientific capabilities of the system and guiding the definition of future observing strategies.

Short-range correlated (SRC) nucleon pairs - caused by brief, high-momentum interactions between two nucleons - are dominated by neutron-proton pairs with large relative and smaller center-of-mass momenta. However, the underlying dynamics that determines which nucleons form such pairs remains uncertain. Previous measurements showed that proton pairing probabilities increased strongly with nuclear asymmetry N/Z, but could not rule out an increase with nuclear mass A. We measured high-missing-momentum protons knocked out in electron scattering from selected nuclei with a range of shell configurations, A, and N/Z, including 9Be, 10B, 11B, 12C, 40Ca, 48Ca, 54Fe and 197Au. Unexpectedly, we found that while the pairing probability increased with A, the slope of the increase was much greater from Be to C and from 40Ca to Fe, than from Be to Au. This shows the importance of long-range nuclear shell structure on the probability of short-range nucleon pairing.

Direct evidence of exciton condensation in semiconductor heterostructures remains elusive. Here we propose charge transport of doped carriers as a probe of exciton condensation in transition-metal dichalcogenide heterostructures and identify distinct experimental signatures. First, condensation suppresses the phase space for carrier scattering, leading to a reduction in resistivity, that provides a general diagnostic of exciton condensation. Second, in heterostructures with a tunable solid-state Feshbach resonance, condensate-induced hybridization between doped carriers and trion bound states qualitatively modifies transport. In particular, near resonance, this hybridization yields a negative effective mass and a corresponding sign reversal of the Hall resistivity. These results establish charge transport as a promising route for detecting and characterizing exciton condensation in semiconductor heterostructures.

The Euclid Quick Data Release (Q1) is a powerful dataset to study active galactic nuclei (AGN) and their host galaxies. Deriving their physical properties through multi-component spectral energy distribution (SED) fitting is a challenging task for AGN, but it is greatly aided by the Euclid near-infrared photometry. Here we present a new method to quantify the reliability of SED-derived parameters, such as AGN bolometric and monochromatic luminosities, host's stellar mass $M_\star$, star-formation rate (SFR) and specific star-formation rate (sSFR), by using mock SEDs of AGN built by combining observed SEDs of QSOs and galaxies. We apply this methodology to the ${\sim}1$ million Q1 AGN candidates, constructing a catalogue of AGN and host galaxy properties, alongside their respective reliability values. With a reliability threshold at 0.5, we find 88\% of sources with robust stellar masses and 76\% with reliable AGN luminosities. Moreover, through SED fitting we also measure the AGN fraction $f_{\rm AGN}$ of the total mid-infrared flux and we use its lower-limit to select AGN. A $f_{\rm AGN, \, low} > 0.075$ threshold yields 85\% completeness and purity. Comparable to colour-colour AGN selections, this method has the advantage of being less affected by redshift evolution and exploring fainter magnitudes. Additionally, by comparing the AGN and host galaxy parameters across different identification methods, we find that the probed range in stellar mass and AGN luminosity can be quite different. This highlights the importance of combining different approaches and accounting for their selection biases when studying AGN and their role in galaxy evolution. Finally, for the X-ray detected sample, we present the X-ray to mid-IR luminosity relation, and the correlation between stellar mass and bolometric luminosity as a function of redshift, in good agreement with previous results.

We review the flavor phenomenology of light dark particles, focusing on axion-like particles with sub-GeV masses and generic flavor-violating couplings. Such states can naturally emerge from the spontaneous breaking of generic flavored symmetries, and are motivated by dark matter or the Strong CP Problem, with the QCD axion serving as a paradigmatic example. Light dark particles can be produced in two-body decays of Standard Model particles, giving rise to missing energy signals that can not only be observed in high-precision flavor experiments, but also be probed in core-collapse supernovae and the cosmic microwave background. These decays are controlled by dimension-five operators, which makes dedicated laboratory searches sensitive to very large UV scales up to $10^{12} {\rm GeV}$ and thus highly complementary to astrophysical and cosmological probes. We provide a comprehensive survey of the resulting limits and prospects across all relevant channels, highlighting the central role of flavor physics in exploring the landscape of light dark matter.

In this note, we introduce the \emph{partial order decomposition number} of a digraph $D$, denoted $pod(D)$, defined as the minimum integer $k$ such that $A(D)=A(P_1)\cup\cdots\cup A(P_k)$, where $P_1,\ldots,P_k$ are partial orders on $V(D)$. We prove that $\dic(D)\le \diomega(D)^{pod(D)}$ for every digraph $D$. In particular, every class of digraphs with bounded $pod$ is polynomially $\dic$-bounded. We apply this to tournaments, showing that if $\mathcal C$ is a class of tournaments with bounded dichromatic number, then the closure of $\mathcal C$ under substitution is polynomially $\dic$-bounded, thereby making progress on a question of Aubian, Charbit, Lopes, and the first author. As further applications of $pod$, we prove that poset tournaments of bounded dimension are $\dic$-bounded, derive polynomial lower bounds on the directed clique number of an explicit family of tournaments, thereby answering a conjecture of Gutowski and Rams, and show that tournaments with bounded $pod$ have bounded domination number.

We present the first installment of CMBolic: a suite of symbolic cosmic microwave background (CMB) emulators. In this instance, we emulate the CMB lensing potential power spectrum $C_\ell^{ϕϕ}$ for the widely used extended $Λ$CDM model which simultaneously includes massive neutrinos and evolving dark energy modelled using the Chevallier-Polarski-Linder (CPL) parameterization. We achieve comparable precision to existing neural network emulators, with the added benefit of simpler handling as our emulators are analytic functions of the model parameters and multipole $\ell$. On independent validation spectra evaluated in the range $2\leq \ell \leq 5500$, CMBolic achieves mean absolute fractional errors of $0.27\%$ in the $Λ$CDM subspace and $0.32\%$ across the full extended parameter space. This emulation error is well below even the most optimistic noise forecasts from CMB Stage 4 experiments. We apply CMBolic to cosmological parameter estimation with Bayesian inference using the lensing-only likelihoods from ACT DR6 and Planck. We show excellent agreement between the posteriors obtained by CMBolic and the Boltzmann code CLASS. This demonstrates the practical use of CMBolic on cosmological parameter estimation, reducing the runtime from 2 weeks to under 3 minutes.

Using the open quantum systems formalism, we study the propagation of a quark-antiquark pair propagating through a dense QCD plasma of size $L$ and transverse momentum broadening transport coefficient $\hat q$, and we derive the Lindblad evolution equation for the density matrix of the system. We focus on the boosted regime where the opening angle $θ_{q\bar q}$ of this effective colour dipole satisfies $θ_{q\bar q}\ll 1$. In the correlation limit where the quark-antiquark relative transverse momentum $p_\perp$ is much larger than the imbalance $q_\perp$ as well as the medium typical transverse momentum scale $Q_s=\sqrt{\hat q L}$, we demonstrate that the resulting Wigner distribution displays quasi factorisation between a hard factor describing the hard splitting producing the $q\bar q$ pair and the transverse momentum imbalance $q_\perp$-distribution encoding the broadening induced by the medium. The factorisation is violated by a "colour decoherence" factor that controls the $θ_{q\bar q}$ dependence of the $q_\perp$-distribution through the ratio $θ_{q\bar q}/θ_c$, with $θ_c \sim (\hat q L^3)^{-1/2}$. The open quantum systems approach enables us to clarify the role of this critical angle $θ_c$ and its associated critical time $t_c$ in the genuine quantum decoherence of the density matrix in colour and kinematic space: in particular, $t_c$ controls both the suppression of the off-diagonal elements of the density matrix in colour space and the transition between singlet and octet states. We find, however, that colour decoherence sets in earlier than the full decoherence of the density matrix, thereby marking the onset of classical behaviour in the system. Finally, we investigate the corrections beyond the quasi-factorised picture due to the quantum diffusion term in $p_\perp$ of the Lindblad equation and we find that these corrections are mild.

Previous studies have shown that cascade production in the thick target of the SHiP experiment may substantially enhance the number of light long-lived particles (LLPs) decaying in the fiducial volume. However, cascade-produced LLPs are typically soft, so daughter-level acceptance and reconstruction effects can strongly suppress the observable event rate. We quantify this suppression for two representative cases: photophilic axion-like particles produced in electromagnetic cascades, and heavy neutral leptons produced in decays of secondary kaons. We combine a semi-analytic event-rate calculation with a detector-level study of ALP reconstruction in the electromagnetic calorimeter. For the nominal SHiP detector design, cascade ALPs give at most a moderate enhancement over primary production, and only at the lightest masses; at higher masses, the cascade contribution becomes subdominant or negligible. For HNLs from secondary kaons, the cascade contribution is already subdominant after imposing daughter-level geometric acceptance. We also identify possible ways to recover part of the cascade event rate, including relaxed event-selection criteria and an active-target subdetector.

Neutron stars are very dense media in which axions may be produced. This has been used to set limit on the QCD axion mass, usually under the assumption that only neutrons, protons, electrons, and muons appear in the star core. Given the extreme conditions reached within neutron stars, it is reasonable to consider that other particles, such as hyperons and $Δ$ resonances, may exist on-shell. In this work, we study how the limit on the mass of QCD axions, namely KSVZ and DFSZ invisible axions, is altered when different equations of state are used, allowing for heavier particles to appear in the neutron star core. We find that this dependence is in general mild and thus reinforces the reliability of the known limit. Additionally, in the DFSZ scenario, it may drive the limit within the sensitivity window for IAXO. This would allow this experiment to discern the composition of neutron star cores if an axion were to be observed within that window.

Recent observations reveal a universal X-ray to H$α$ surface-brightness ratio, ${\rm SB}_{\rm X}/{\rm SB}_{\rm Hα}\sim 3$, in galactic winds, ram-pressure stripped tails, and cluster filaments. This is surprising because H$α$ traces cold ($\sim 10^4$ K) gas while X-rays trace much hotter ($\sim 10^{6}$--$10^{7}$ K) gas. Plane-parallel mixing-layer models do not recover this ratio, and can be off by orders of magnitude. Motivated by recent work showing that geometry controls the temperature PDF of multiphase gas (Chen & Oh 2026), we run 3D wind-tunnel simulations in the high density contrast ($χ\sim 10^3$) regime. In this limit, the cold phase shatters into many small H$α$-emitting clumps, while X-ray-emitting gas forms a volume-filling cocoon around them. After smoothing on the tail-width scale, the measured surface-brightness ratio converges to the observed value, which can be understood theoretically. The H$α$ luminosity fraction is set by atomic physics, whereas the X-ray luminosity fraction is set by the residence time of gas in the X-ray-emitting band. This residence time is much shorter than the cooling time at X-ray temperatures, but scales roughly inversely with pressure, suggesting that it is tied to the cooling time at a lower-temperature outlet of the mixing cascade. This framework naturally explains why the observed ratio is order unity, and robust to changes in gas pressure.

Magnetic Skyrmions are planar solitons stabilized by the Dzyaloshinskii-Moriya interaction (DMI) and realized in chiral magnets. We study their natural three-dimensional generalization: a sigma model from $\mathbb{R}^3$ to $S^3$ with a four-component magnetization vector, stabilized by a one-derivative term which is a generalized DMI. We utilize two SO(3)-invariant generalized DMIs discovered recently: an "$α$-term" supporting a spherically symmetric hedgehog Skyrmion and a "$β$-term" supporting an axially symmetric Skyrmion that splits into two half-Skyrmions connected by a magnetic string of negative tension, a phenomenon we call "anti-confinement". We derive a cubic-lattice discretization that reproduces both continuum theories at long wavelengths and use Monte Carlo simulations to map the finite-temperature phase diagram. We identify spin-spiral, magnetic-string-lattice, Skyrmion-lattice, and antiSkyrmion-lattice phases, as well as a mixed-topology regime with fractional $S^3$ charges localized at string bends. We find, for the first time in the literature to the best of our knowledge, a toroidal (anti-)soliton of unit charge. Our results establish a theoretical and computational framework for three-dimensional topological magnetic textures in systems whose order-parameter manifold is $S^3$.

We point out that ultra-peripheral Au+Au collision data collected at the Relativistic Heavy Ion Collider, operational during 2000-2026, can be used to search for axion-like particles coupled to photons via the resonant process $γγ\to a \to γγ$. Exploiting the $Z^4$ enhancement of the two-photon luminosity in heavy-ion collisions and the low photon energy thresholds achievable at RHIC, we simulate signal and background processes, the latter dominated by light-by-light scattering, hadronic resonance production, and misidentified $e^+e^-$ pairs, and estimate upper limits on the ALP-photon coupling $g_{aγγ}$ assuming $1.9~\text{nb}^{-1}$ of existing data collected by the PHENIX experiment. We find sensitivity to ALP masses in the range $2~\text{GeV} \lesssim m_a \lesssim 5~\text{GeV}$ with couplings $g_{aγγ} \gtrsim 4\times 10^{-4}~\text{GeV}^{-1}$, probing previously unexplored regions of parameter space. Access to larger luminosity datasets could substantially extend the sensitivity of this search, motivating a dedicated analysis of ultra-peripheral collision data collected at RHIC by PHENIX as well as other experiments.

Wilson lines are fundamental probes of gauge theories. We show that, in sufficiently rich representations, they support a large new class of operator insertions. For half-BPS lines in $\mathcal{N}=4$ SYM many of these operators have the quantum numbers of the displacement supermultiplet. Their dimension-one superprimaries define natural deformations of the defect theory. By analyzing the associated beta functions, and relating them to specific OPE coefficients, we show that the deformations are marginally relevant. We support our finding with a weak-coupling computation of the four-point function of these operators for any gauge group and representation.

We provide a complete classification of twisting supercharges for type IIA supergravity in a flat background by determining the orbits of square-zero elements in the (1,1) super Poincaré algebra in ten dimensions under the action of the Spin group. Together with recent results for type IIB, this completes the description of the moduli space of twists for type II supergravities. Further, we discuss the origins of these twists as dimensional reductions from eleven dimensions. Using the pure spinor formulation of the superstring, we relate the target space classification to worldsheet twists and identify supergravity counterparts of mixed A/B models. Finally, we work out the action of T-duality that relates twists of type IIA with those of type IIB.

Metastability in many-body quantum systems and its associated exponentially-long timescales have been the subject of considerable recent interest. Here, we focus on a class of driven-dissipative many-body open quantum systems described by a Lindbladian having hidden time-reversal symmetry (a form of quantum detailed balance). Examples include boundary-driven interacting spin chains, bosonic lattice models and driven-dissipative collective spin models. We suggest that for such systems, slow timescales in the vicinity of a dissipative first-order phase transition can be analytically predicted using a special purification of the non-equilibrium steady state. We show the accuracy of our conjecture through detailed studies of a dissipative transverse-field Ising model with collective and local decay, and a driven-dissipative nonlinear cavity model. Our results allow quantitative insights into metastability and slow dynamics for a range of systems, including cases where semiclassical or path-integral instanton approaches are intractable.

Producing high-fidelity magic states using the smallest possible amount of physical qubits and operations stands as a very important challenge to achieve fault-tolerant quantum computation at scale. Besides emerging proposals for alternative methods such as cultivation, magic state distillation remains essential for achieving very low error rates. Known distillation protocols are usually built through quantum codes derived from triorthogonal matrices. Here, exploiting the specific noise structure present in magic state distillation protocols, we show that classical error-correcting codes offer a simpler framework for deriving these protocols. This formulation is particularly well suited to systematic numerical and analytical studies of distillation protocols involving a fixed number of qubits. Specifically, we use a SAT solver to derive a series of no-go theorems that relate key figures of merit, including the number of qubits, the protocol depth, the factory distance, and the prefactor in the output error rate. For instance, we prove that any $T$-to-$T$ state distillation protocol using fewer than eight qubits can detect at most three errors, while any $T$-to-$\mathrm{CC}Z$ state distillation protocol using fewer than eight qubits can detect at most two errors. Our results also include new distillation protocols with the smallest number of qubits for a given distance in the literature, namely distance 4 and 5 $T$-to-$T$ state protocols supported on 10 and 11 qubits, as well as distance 3 and 4 $T$-to-$\mathrm{CC}Z$ state distillation protocols supported on 9 and 10 qubits.

The Galactic Center (GC) contains a collection of filaments that are typically tens of parsecs in length, illuminated by synchrotron radiation from cosmic rays (CR). The origin of these nonthermal filaments (NTFs) is unclear. We aim to distinguish two injection mechanisms: the first mechanism posits that NTFs are fueled either by jets from pulsar wind nebulae and are lepton-dominated; the second mechanism posits that NTFs are fueled by accelerated particles from interstellar shocks and are proton-dominated. We explore these mechanisms using the magnetohydrodynamics (MHD) code Athena++, modified to account for radiative and collisional losses, to simulate CR propagation with lepton and proton CR species. We vary parameters such as magnetic field strength, plasma density, and the CR diffusion coefficient to determine how the range of conditions present in the GC can affect CRs' propagation, heating, plasma flow, and the observed synchrotron emission. We find few observable differences between the proton- and lepton-dominated cases, but comparing the models with observed filament properties motivates consideration of a third formation mechanism: the generation of NTFs arise from intermittent structures in Galactic Center turbulence.

A Windows driver exposes its kernel through I/O control (IOCTL) codes, and a single unchecked length on the buffer behind one turns an unprivileged call into a kernel write. The research community has strong scanners for this surface and a curated list of known-bad drivers, but no map of the surface itself. We build that map. The Windows IOCTL Census is a queryable database of the control-code dispatch surface of 27,087 signed Windows drivers, recovered by one deterministic, architecture-neutral pass with no symbolic execution. Reading a lifted intermediate representation instead of running a symbolic engine lets it recover a dispatch surface for 80% of the corpus across x86 and x64, including the 32-bit half existing scanners abort on. On the 64-bit lane it adds handler reachability, taint, and the call graph. An LLM ranks the reachable handlers for triage. We release the census as a public dataset of tens of millions of rows: 27,087 binaries, 3.1M decoded control codes, 8.18M functions, and 15.95M call edges.

We investigate whether current astrophysical data constrain not only the high density continuation of the neutron star equation of state, but also the low density matching procedure itself. To this end, we compare two low density branches propagated through a common high density extension and confront them with direct NICER mass-radius posteriors, a lower bound on the maximum mass, and an effective constraint on $Λ_{1.4}$. We find that the observable predictions of the two branches remain strongly overlapping, while the NICER-informed posterior still induces a nontrivial constraint on the matching parameters. Current data therefore constrain primarily the shared continuation above $n_1$, but also indirectly restrict the low-density matching sector.

We present a new, highly homogeneous quasar sample with X-ray and UV observations optimized to reliably estimate distances via the non-linear X-ray-to-UV relation. Cross-matching the Sloan Digital Sky Survey DR16 quasar catalog with the XMM-Newton serendipitous catalogue (4XMM--DR14), we employ strict selection criteria to build a robust sample: (1) UV and (2) X-ray colour constraints to avoid, respectively, extinction and absorption; (3) removal of broad absorption line and radio-bright quasars; (4) exclusion of sources at z<0.7 to prevent galactic UV contamination; and (5) rejection of sources with shallow X-ray observations. The latter step, closely related to the Eddington bias, is critical because SDSS data are generally deeper than X-ray data for typical quasar spectral energy distributions: ignoring such a discrepancy introduces a spurious redshift dependence in the X-ray-to-UV relation parameters. Our final sample contains about 2,000 quasars at z=0.7--5. We demonstrate that the X-ray-to-UV relation is constant across this redshift range, with a mean slope of 0.58\,$\pm$\,0.01 and a dispersion of 0.15 dex. Our findings confirm the intrinsic stability of this relation over cosmic time, emphasizing that both homogeneity and robust Eddington bias corrections are vital for flux-limited samples. In fact, the impact of the preferential detection of X-ray brighter-than-average sources near the effective sensitivity limits significantly grows with redshift. Any resulting evolutionary trend in the X-ray-to-UV relation, especially in the form a slope flattening, is therefore either a largely spurious effect, or the result of mixing populations of quasars with intrinsically different spectral properties.

The ensemble Kalman filter (EnKF), originally developed in the geophysical sciences, is now widely used for state and parameter estimation problems in various domains of application. It may be viewed as a robust, cheap-to-implement, alternative to the optimal, Bayesian, filter. However, despite its empirical successes, theoretical understanding of its properties, in relation to the optimal filter, is in its infancy. In this paper we contribute novel theoretical understanding, studying the behavior of the mean-field limit of the ensemble Kalman filter (MFEnKF) in the linear setting. In this setting the MFEnKF coincides with the Kalman filter itself, for Gaussian initial data. We study the MFEnKF with general, non-Gaussian, initial data. Under the assumptions of nondegeneracy of the signal noise and detectability of the signal-observation pair we make three specific contributions. First, we derive the MFEnKF via a variational approximation of the Bayes update using a covariance-weighted optimal transport metric. Secondly, we use this metric to show that the MFEnKF is a strict contraction towards the subspace of Gaussian measures, giving explicit geometric rates in terms of the covariances, in both discrete and continuous time. Finally, in the continuous-time setting, we deduce a stable form of almost sure accuracy of the MFEnKF: given any pair of (nondegenerate) initial data for the MFEnKF and the optimal filter, any uniformly continuous moments of the two filtering distributions coincide in large time, for almost every observation path.

Accurate and reliable attitude determination (AD) is essential for unmanned vehicles operating in Global Navigation Satellite System (GNSS)-denied environments. Short-range wireless arrays can provide direction-of-arrival (DoA) measurements from multiple anchors, enabling AD by aligning corresponding direction vectors (DVs) expressed in the body and navigation frames. In short-range scenarios, navigation-frame DVs inherit non-negligible uncertainty induced by anchor/vehicle position errors in addition to DoA-induced errors in body-frame DVs. Moreover, due to projection and unit-norm normalization, the DV errors are generally anisotropic, which motivates a total least squares (TLS) viewpoint. This paper identifies the key modeling distinction in short-range AD, develops a TLS-consistent formulation based on the total DV error and solves the resulting covariance-weighted orthogonal Procrustes problem via a manifold Gauss--Newton method. To retain the efficiency and numerical robustness of the closed-form weighted Wahba solution, we further propose Hessian-matching based scalar weighting strategies that approximate the Hessian of Wahba formulation to the TLS formulation, including a full-attitude strategy for overall accuracy and a direction-of-interest (DOI) strategy for prioritizing a selected attitude component. Finally, we incorporate IMU-derived gravity as an additional DV pair for static initialization, leading to extended Wahba and extended TLS formulations. Simulation results demonstrate that the proposed Hessian-matching weighting improves accuracy and robustness compared with existing baselines, and that gravity-DV augmentation further reduces attitude errors and improves solution availability under limited anchor availability.

This paper studies specification testing in dynamic linear models in the presence of omitted variables. The null hypothesis of interest is weak exogeneity: shocks have zero conditional expectation given their own past and the past of omitted variables. Existing tests based on quadratic forms of serial cross-correlations suffer from size distortions because their variance incorporates symmetric dependence in both directions, including causality from past shocks to present omitted variables (inverse causality). This paper proposes an asymmetric Portmanteau test that isolates violations of weak exogeneity from inverse causality, is asymptotically normal under the null, and does not require a parametric specification of the joint dynamics. An empirical application examines the Economic Policy Uncertainty shock series and rejects its weak exogeneity. Addressing this failure by controlling for omitted variables changes the estimated inflation response from negative to positive, suggesting a supply-side shock interpretation.