Papers

Prefix caching is critical for efficient large language model (LLM) serving, particularly for agentic workloads that repeatedly invoke the model with a growing conversation and tool-use history. By reusing the key-value (KV) states of previously processed prefixes, prefix caching avoids redundant prefill computation. Its effectiveness, however, depends on retaining a sufficiently large set of KV cache states. Provisioning enough cache to preserve all historical KV states is prohibitively expensive and often unnecessary, whereas insufficient capacity can substantially degrade the cache hit rate. Determining the KV cache working set, defined as the minimum cache capacity required to achieve a target hit rate, is therefore essential for efficient cache provisioning and system design. We present KVSET, an online analyzer that estimates the KV cache working set of LLM serving workloads. KVSET uses the Mattson stack algorithm to efficiently estimate cache hit rates across a wide range of cache capacities. For each KV cache page, KVSET computes its LRU stack distance and compares it with the page number of each candidate capacity. This comparison determines whether the page would be a hit at each capacity without independently simulating every capacity configuration. KVSET therefore substantially reduces the computational and memory overhead of conventional capacity-by-capacity simulation and makes online working-set analysis practical. KVSET further determines the minimum cache capacity based on the maximum LRU depth among the prefix pages required to achieve the target hit rate. We validate KVSET using traces collected from production LLM workloads and show that its estimates closely match measurements from real cache deployments. The open-source implementation supports both online request processing and offline trace replay.

This paper advocates category theory as a practical framework for structuring and improving rein- forcement learning in high-dimensional, partially observable environments. We model symmetries between environmental states by partitioning the state space into equivalence classes induced by sym- metry orbits, and organise each such class as a groupoid with a designated canonical representative. This allows the agent to share what it learns across many similar environmental states simultaneously, rather than treating every orientation or position as an entirely new problem. Learning is thus carried out on a symmetry-reduced state space with each orbit represented once, preserving structure while eliminating redundancy and improving sample efficiency. We implement this framework within standard reinforcement learning pipelines and evaluate two different approaches on partially observable benchmarks, demonstrating that orbit-based partitioning yields consistent performance improvements in environments exhibiting latent symmetry. Beyond these empirical results, our approach illustrates how categorical structure provides a principled bridge between abstract reinforcement learning formulations and their computational application, thereby establishing a pathway toward more structured and scalable learning systems.

Seasonal hydropower operation requires release decisions that account for uncertain future inflows and the opportunity cost of stored water. We assess the Modelling and Optimisation of Stochastic Seasonal Hydropower Storage (MOSSHOOS) framework for the Manantali reservoir using 660 monthly energy-inflow observations from 1961–2015. The framework combines multiscale inflow representation, blocked resampling and scenario reduction with stochastic dual dynamic programming, and uses a seasonal square-root diffusion and Hamilton–Jacobi–Bellman formulation as a continuous-time interpretation of marginal water value. Scenario reduction improves the energy score by 47.6% relative to independent monthly sampling, but the marginal distribution is rejected by a Kolmogorov–Smirnov test and the lower-tail p10 error remains 103.9%, indicating inadequate dry-tail calibration. Consistent with that limitation, 50 paired Monte Carlo evaluations show no detectable cost or deficit advantage of the stochastic policy over its deterministic-equivalent counterpart at the archived operating configuration. As an independent benchmark, the repaired open-source plan4res seasonal storage valuation chain yields a dominant marginal value of 81.66~USD/MWh, within 3% of the MOSSHOOS mean of 79.25~USD/MWh. These results support a coherent stochastic inflow-to-water-value methodology while showing that hydrological tail representation, rather than economic scaling, is the principal limitation to address before operational application.

A reinforcement learning policy that satisfies its temporal-logic specification has passed a test, not an assurance argument. The clause that matters to a reviewer may never have mattered to training: it may have been avoided entirely, forced by the environment regardless of what the policy learned, or redundant next to the ordinary task reward. Satisfaction probability and task return cannot tell any of this apart. This paper introduces an audit layer that can. It measures whether a specification clause was actually exercised, whether that role was forced or chosen, and whether the obvious way to test causation, weakening the clause and retraining, is even valid. Often it is not: we prove that comparable weaker/stronger training objectives can share perfect optima under standard acceptance-derived rewards, show related ablation hazards across a large corpus of published specifications, and then show that a properly designed intervention detects the effect it should. Across standard reinforcement learning benchmarks and published external artifacts, the audit layer separates six regimes that a single satisfaction number collapses into one. A policy that satisfies its specification has answered whether. This paper asks why.

Many-body environments are conventionally incorporated into quantum dynamics as heat baths, which induce a Lamb shift and quantum dissipation. In the traditional picture, such environmental coupling is expected to induce suppression and broadening of spectral signals. In this letter, we investigate high-order harmonic generation (HHG) in a two-level system coupled to a heat bath via dipole-dipole interactions using the Lindblad master equation. It is found that the environmental effect does not simply suppress the harmonic efficiency. Instead, when the cutoff frequency of the coupling spectral density exceeds the energy-level spacing, the environment can actually enhance the harmonic yield. Further analysis reveals that this enhancement originates from intense level fluctuations induced by the Lamb shift. Meanwhile, the environmental influence exhibits a distinct dependence on harmonic order, manifesting clearly different behaviors for lower and higher harmonics. Our results challenge the common relaxation-time picture of uniform damping and establish a microscopic, order-dependent mechanism for environmental control of HHG, with direct implications for solid-state attosecond spectroscopy and quantum material engineering.

In reinforcement learning, more accurate policy evaluation has been frequently shown to improve downstream policy optimization performance. For tractability, most value estimators rely on $n$-step truncation and thereby omit the long-horizon behavior of the transition dynamics. Methods that instead capture this behavior through the global structure of the dynamics can converge faster, but their computational and memory requirements limit their scalability to large or continuous state spaces. To reconcile these limitations, we introduce Limiting-Kernel Q($\lambda$) (LKQL), an off-policy value estimator that uses the limiting kernel to approximate the long-horizon terms omitted by truncation. LKQL has the same order of complexity as $n$-step estimators and integrates directly into both on- and off-policy actor-critic algorithms. We prove that, under aperiodicity and in the near-on-policy regime, the operator underlying LKQL improves the policy evaluation convergence rate over its truncated counterpart for sufficiently large $n$, and that LKQL itself converges almost surely to the optimal values in finite Markov decision processes (MDPs) under a fixed behavior policy. On the MuJoCo continuous-control benchmark, we show that LKQL improves over $n$-step baselines in most settings, particularly on long-horizon tasks.

We study the quantized six-vertex (Q6V) model on arbitrary admissible diagrams at $q=\varepsilon$, where $\varepsilon$ is a primitive $N$th root of unity with $N$ odd. For the commuting layer transfer matrices $\mathbb{T}(y)$ with mixed boundary conditions, we establish the quantum Frobenius property \[ \mathbb{T}(y)\mathbb{T}(\varepsilon y)\cdots \mathbb{T}(\varepsilon^{N-1}y) = \mathscr{P}_N(y^N)\mathbb{I}, \] where $\mathscr{P}_N$ is a scalar Laurent polynomial obtained directly from the monomial expansion of $\mathbb{T}(y)$. Our proof exploits the underlying three-dimensional integrability, in particular the tetrahedron equations and their consequences. The quantum Frobenius relation essentially determines the joint eigenvalues of the associated Hamiltonians up to multiplicities and shows that they take a free-parafermion form, thereby extending such structures from previously known one-dimensional examples to a broad class of genuinely two-dimensional quantized six-vertex models. The construction also admits independent local gauge choices and recovers the long-range-interacting $\tau_{\; 2}$ model, which includes the original free-parafermion model as a special case. We also apply the same framework to the relativistic quantum Toda chain at odd roots of unity and obtain the free-parafermion spectra for the relevant Q6V transfer matrices with fixed boundary states.

We propose the Memory-Efficient Neural Operator (MENO) as a high-performance PDE neural solver based on the Manifold Function Encoder (MFE). MENO features three primary advantages: (1) MENO has a significantly smaller memory footprint and much faster training speed than other popular architectures, with the memory footprint being independent of the data resolution, and therefore holds the potential for scaling up to large-scale models. (2) MENO can accept PDE inputs of arbitrary form, including arbitrary geometric domains and arbitrary discretizations. In particular, it is capable of handling cross-geometry scenarios, i.e., where the input functions and the output solutions are defined on different manifolds. (3) MENO exhibits strong generalization capability, and achieves the best accuracy on most of the benchmarks we tested, compared with the results reported in the literature. The code is available on GitHub at https://github.com/jpzxshi/MENO, and all numerical examples in this paper can be run with a single command to reproduce the reported results.

We study lattice-counting functions on compact integral-integral affine manifolds with corners, which generalize integral unimodular polytopes. We prove analogues of Ehrhart's theorem and Ehrhart-Macdonald reciprocity in this setting.

Shell structure and spontaneous decay of $Z = 124$ isotopes are systematically investigated using the deformed relativistic Hartree-Bogoliubov theory in continuum (DRHBc) with the PC-PK1 density functional. By accounting for deformation and continuum effects, the neutron drip line is predicted to be located at $N = 320$, and eight potential multi-neutron emitters beyond the neutron drip line are identified. The results, compared with the spherically constrained case, highlight the critical role of deformation effects in determining the ground-state properties of superheavy nuclei. In addition, potential neutron shell closures emerge at $N = 258$ and $350$, alongside subshell closures at $N = 232$ and $320$. The results disfavor $Z = 124$ as a possible proton magic number and suggest $Z = 120$ and $138$ as more favorable candidates for proton shell closures in the superheavy region. Finally, the competition between $\alpha$ decay, $\beta$ decay, and spontaneous fission is analyzed by calculating half-lives within various semi-empirical formulas. The results indicate that $\alpha$ decay is the predominant mode on the proton-rich side, whereas spontaneous fission and $\beta^-$ decay gradually become dominant with increasing neutron number.

Let $G$ be a finite group admitting a rational fixed-point-free (FPF) representation $V$, with a $G$-invariant full-rank lattice $L\subset V$. We study the induced $G$-action on the complex torus $V_{\mathbb{C}}/L$ and the quotient orbifold $V_{\mathbb{C}}/(G\ltimes L)$. We prove that every FPF group admits a unique irreducible rational FPF representation. We derive closed-form formulas for three core geometric invariants: the number of non-free points of the torus action, the number of singular points of the quotient orbifold, and its orbifold Euler characteristic.

Newton-Schulz (NS) iteration has recently been used in the Muon optimizer to transform update matrices during the training of large language models. Motivated by its spectral effect, we investigate applying NS directly to Transformer attention representations. We introduce Newton-Schulz Attention (NS-Attn.), a parameter-free transformation applied to the output of each attention head. Each head output is arranged as a feature-by-token matrix and normalized by its Frobenius norm. We then apply a finite NS polynomial step and restore the original norm. The objective is to reduce spectral concentration and increase effective rank before standard head merging and output projection. Across ViT and Swin on CIFAR-10 and CIFAR-100, NS-Attn. improves final-epoch accuracy in all 12 matched-seed comparisons, with mean gains of 0.25–0.83 percentage points. ViT ablations show higher mean accuracy with one iteration than with two. Spectral analysis further shows reduced leading-eigenvalue concentration and increased effective rank. These gains incur additional inference latency.

Vision-Language-Action (VLA) policies often rely strongly on the camera viewpoints seen during training, causing substantial performance degradation when deployed from unseen perspectives. Collecting demonstrations from sufficiently diverse physical viewpoints is expensive and still provides only sparse coverage of the viewpoint space. We introduce InfiNoVA, a data-augmentation framework that converts synchronized multi-camera demonstrations into a dense distribution of geometrically consistent training views. InfiNoVA reconstructs each manipulation trajectory as a time-varying 3D Gaussian representation and renders novel observations from sampled camera poses while preserving the original state-action correspondence. This explicit scene representation improves frame-level fidelity and temporal consistency while reducing task-critical hallucinations observed in generative novel-view synthesis. Across four real-world manipulation tasks, policies trained with InfiNoVA achieve 5.4x higher average success under unseen randomized viewpoints than both VISTA-based augmentation and the unaugmented policy. InfiNoVA further achieves 1.7x higher success than training directly on all five physical camera views. These results show that dense, geometrically grounded viewpoint augmentation provides a practical route toward camera-robust robot policies without modifying the underlying policy architecture.

As extremely large antenna arrays (ELAAs) become central to next-generation wireless systems, the transition into the near-field propagation regime enables the exploitation of spherical wavefronts for radial-domain beamfocusing. This capability is pivotal for emerging applications requiring precise spatial isolation, such as Space Division Multiple Access (SDMA), hierarchical localization and advanced sensing. However, fully realizing these technologies requires specific design rules to dimension multi-user systems without relying on computational expensive full-wave simulations. To address the gap in modeling contiguous focal regions with controllable radial resolution, this paper expands the Angular Spectrum Representation (ASR) approach to propose a comprehensive analytical framework. Through the introduction of a tunable inter-beam overlap parameter $\rho$, we derive closed-form expressions to synthesize multiple focal regions, providing the flexibility to tailor their radial resolution. Furthermore, the resolution capabilities are analyzed to characterize the interplay between key operational variables, such as the transmitter size, beam radius and operation frequency. System-level assessment of per-user and sum-rate spectral efficiencies across varying signal-to-noise (SNR) regimes reveals how the inter-beam overlap dictates a fundamental trade-off between user capacity and inter-user interference, delivering design guidelines for future near-field communications.

This paper estimates the causal effect of drug consumption room (DCR) openings on drug-related hospitalizations in Switzerland. I exploit the staggered introduction of DCRs across Swiss cities between 1998 and 2022, using individual-level hospital records and a difference-in-differences estimator. DCR openings reduce drug-related hospitalization rates by roughly 35% within 5 km of a facility. The reduction is most pronounced for fatal outcomes, which fall by 35% while narcotics poisoning hospitalizations rise by a comparable amount, consistent with DCRs converting potentially fatal overdoses into non-fatal hospitalizations. Mental and behavioral disorder hospitalizations decline by 32% while estimates for Hepatitis and HIV are imprecise. The findings support the role of DCRs as an effective harm reduction strategy.

A subset $S \subseteq \mathbb{F}_2^n$ is called a Sidon set if no four distinct points of $S$ have zero sum. It is shown that if $n \geq 7$ and $n \equiv 3 \mod 4$, then $|S|\leq 2^{\frac{n+1}{2}}-3$. As a consequence, for any even $t \geq 4$, there does not exist a binary linear $[2^t-3,2^t-2t-2,5]$-code, strengthening a nonexistence result of Brouwer and Tolhuizen from 1993. As a second consequence, we also show that for an even integer $n \geq 4$, every almost perfect nonlinear (APN) function $F \colon \mathbb{F}_2^n \to \mathbb{F}_2^n$ has nonlinearity at least $3$.

Let $G$ be the group $\mathbb{R}_+\ltimes \mathbb{R}^n$ endowed with Riemannian symmetric space metric $d$ and the right Haar measure $\mathrm{d} \rho$ which is of $ax+b$ type, and $L$ be the positive definite distinguished left-invariant Laplacian on $G$. Let $u=u(t,\cdot)$ be the solution to $u_{tt}+Lu=0$ with initial conditions $u|_{t=0}=f$ and $u_t|_{t=0}=g$. In this article we show that for a fixed $t \in\mathbb{R}\setminus\{0\}$, \begin{align*} \|u(t,\cdot)\|_{L^{1,\infty}(G,\mathrm{d}\rho)}\leq C\ (1+|t|)\big(\|(\operatorname{Id}+L)^{\alpha_0/2}f\|_{L^1(G,\mathrm{d}\rho)} +\|(\operatorname{Id}+L)^{\alpha_1/2}g\|_{L^1(G,\mathrm{d}\rho)}\big) \end{align*} if and only if $\alpha_0\geq {n/2}$ and $\alpha_1\geq n/2-1$. Moreover, the polynomial-in-time growth $1+|t|$ in the above estimate is best possible.

In neuropsychiatry, the primary goal is often not only to decode brain activity but to change it, for example to lessen a negative affective bias or an overly salient memory. Motivated by control theory, we develop an AI-driven neural-surrogate framework that proposes candidate representational changes and tests their predicted perceptual effects from snapshots of stimulus-evoked fMRI activity, without physical stimulation. The framework combines fMRI decoding, deep generative modeling, and constrained latent-space steering. Valence and memorability are used only as worked examples. Using more than 36,000 image-fMRI observations from four deeply sampled Natural Scenes Dataset participants, subject-specific models recovered coarse generative structure from visually responsive cortex (two-way identification, 0.79-0.88; chance, 0.5). Graded perturbations were reconstructed as images and evaluated with automated scorers and human ratings from 7,200 trials by 18 participants. In the primary VDVAE model, valence shifted from -0.61 to +1.03 SD and memorability from -1.34 to +1.45 SD; a later Versatile Diffusion refinement reduced or altered these effects. Across five perturbation levels, human valence ratings moved in the predicted direction under the linear time-correction model (mean slope, 0.038 SD per unit of alpha; 95 percent CI, 0.003-0.074; positive in 16 of 18 participants). Perceived memorability did not change reliably. Baseline agreement with the automated assessor was suggestive for valence (r = 0.30) and weak for memorability (r = 0.10). Extreme perturbations drifted from the original stimulus, so intended change must be weighed against loss of fidelity. These findings provide a falsifiable upstream method for designing and behaviorally testing candidate representational targets for future neuromodulation in psychiatry, while marking the limits of the present static approximation.

We introduce NeuralSRNF, a novel framework for the statistical shape analysis and generation of genus-zero 3D and 4D objects that undergo nonrigid deformations. Traditional methods rely on complex and computationally expensive nonlinear elastic metrics that measure bending and stretching. Recent advances in elastic shape analysis achieve computational efficiency by mapping input 3D shapes to the space of Square Root Normal Fields (SRNFs) where the L2 metric approximates the partial elastic metric, significantly facilitating the process of computing geodesics and summary statistics. SRNFs, however, are not invertible, and the numerical algorithms used to map SRNFs back to the original space of surfaces remain computationally very expensive and often lead to approximate results. This paper addresses this fundamental SRNF inversion problem using a novel neural representation, termed NeuralSRNF. Unlike the commonly used numerical SRNF, NeuralSRNF is (1) continuous, and thus resolution-agnostic, enabling full functional shape analysis, (2) more accurate, and (3) computationally more efficient as it can compute inverse SRNF maps along a geodesic path in less than 3 s compared to over 10 min for the numerical SRNF. We demonstrate, using various datasets, the utility and efficiency of the proposed NeuralSRNF in multiple elastic 3D and 4D shape analysis tasks such as geodesic computation, deformation transfer, statistical summaries computation, and 3D shape generation. We show that it outperforms competing methods on most evaluated datasets and metrics by a wide margin in both accuracy and computational efficiency. The source code and additional results are available at https://awaisnizamani16.github.io/awais/NeuralSRNF/.

A Central Bank Digital Currency (CBDC) is central-bank money in digital form, held by the public. Leading designs have two limitations: conversion from bank deposits into CBDC can accelerate deposit flight, requiring safeguards, and the CBDC stays outside credit creation and monetary-policy operations. Here we present a CBDC architecture based on grassroots currencies that overcomes these limitations. The architecture has three components: (1) Money: sovereign grassroots coins, which are digital debts of one unit of fiat currency issued by the central bank, constituting a direct CBDC; (2) Credit and Liquidity: non-sovereign grassroots coins, which are digital debts of one unit of the same fiat currency, redeemable at par, that can be issued by any person, natural or legal, thus adding credit; and (3) Interest: grassroots bonds, sovereign and non-sovereign, adding maturity and thus interest, standard banking instruments, and the central bank's instruments of monetary policy. The central bank can therefore lend, absorb liquidity, set its rates and buy and sell securities in the coins and bonds the public holds, choosing the counterparties and terms of its credit operations, and without converting bank deposits into newly issued central bank money on demand. We prove that the arbitrage-free price of any non-sovereign grassroots coin whose issuer redeems it on demand is one unit of the fiat currency. The central bank can choose to deal with any counterparty, not just banks, and we argue that the central bank's interest rates on lending and bonds bound from above and below the corresponding interest rates of its counterparties. Sovereign and non-sovereign grassroots coins and bonds have been implemented and tested on a small scale.

As AI systems are increasingly integrated into professional work, reflection strategies such as cognitive forcing and prompts that foster critical engagement have shown promise in reducing overreliance and improving decision quality. However, these strategies have primarily been evaluated as short-term interventions within single sessions. The next challenge is to assess whether such mechanisms sustain human agency and expertise over time. Drawing on prior work in AI-assisted decision-making, metacognition, and reflective AI engagement, we examine the challenges of designing and evaluating reflective mechanisms for long-term skill sustainability, considering individual differences in how users engage with such support, the organisational conditions under which it is implemented, and the gap between short-term evidence and long-term claims. We introduce open questions for the research community about the conditions under which reflective AI engagement can be sustained in practice.

Tuning PID controllers to satisfy competing objectives - low tracking error, fast settling, limited overshoot, and moderate control effort - is labor-intensive and requires expertise. Genetic algorithms (GAs) offer gradient-free optimization of controller gains against a weighted fitness function, but success depends on meta-level choices: population size, generation budget, gain bounds, and fitness weights. These are usually set by manual trial-and-error or costly bilevel optimization, exposing a tension: GAs excel at dense numerical search, but configuring them needs high-level, context-dependent semantic reasoning. We propose GA-Agent, which decouples these modes. A standard GA handles low-level PID gain optimization. A large language model (LLM) agent operates at the meta-level: it observes completed GA runs, diagnoses gaps versus user control objectives, and proposes updated GA configurations. The architecture uses structured memory, quantitative goal translation, resource-aware termination, and outcome-driven routing. We evaluate GA-Agent on eight control case studies with diverse dynamics (DC motor, inverted pendulum, aircraft pitch, autonomous underwater vehicle, and others). GA-Agent achieves 100% success on all benchmarks, outperforming a Regular GA with fixed hyperparameters in solution quality and sample efficiency. It matches or surpasses a Cascade-GA baseline while reducing function evaluations by one to two orders of magnitude, typically converging in one to three optimization attempts. Sensitivity analysis shows robustness across LLM backbones and memory configurations. A compact memory buffer (size 2-3) and cost-effective models (DeepSeek-V4-Flash at about $0.002 per run) achieve superior performance.

Silicon-based spin qubits in Si/SiGe heterostructures are a leading platform for scalable quantum computing, yet bridging the gap between theoretical computer-aided design(CAD) models and experimental reality remains a significant challenge. Standard simulations often fail to capture critical physical phenomena, such as interface dipoles, parasitic charge accumulation, and the magnetic hysteresis of on-chip micromagnets. In this work, we present a comprehensive, experimentally calibrated 3D simulation pipeline for a 6-dot Si/SiGe device. We refine the semiconductor band alignment and introduce a semi-empirical classical charge screening model to accurately capture the formation of parasitic wells and their suppression of gate lever-arms. Furthermore, we apply the Jiles-Atherton model to account for the hysteresis and pre-magnetization of integrated cobalt micromagnets, successfully reproducing the experimental resonant frequencies across all six qubits. By coupling these calibrated electrostatic and magnetic profiles into a time-dependent rotating wave approximation (RWA) Hamiltonian, we reproduce experimental observables, including microwave power chevrons. This framework provides a robust foundation for predicting device behavior, evaluating microwave line losses, and optimizing future scalable spin qubit architectures prior to fabrication.

We investigate superfluid pairing in a two-dimensional Rashba spin-orbit-coupled Fermi gas subject to spin-selective one-body loss. Within the non-Hermitian mean-field framework, we self- consistently solve the gap and number equations and find that moderate dissipation can significantly enhance the pairing gap, resulting in a pronounced nonmonotonic dependence on the dissipation strength. Dissipation also provides an additional control parameter for driving the system across the BCS-BEC crossover. We further analyze the quasi-particle spectrum and identify two distinct superfluid regimes characterized by one and three exceptional rings, separated by an exceptional spectral transition. Interestingly, dissipation can enhance both pairing channels while simultaneously inducing a momentum-dependent phase twist in the triplet component. These results demonstrate that spin-selective dissipation provides a versatile non-Hermitian control knob for manipulating superfluid pairing, spectral structure, and crossover physics in spin-orbit-coupled quantum gases.

We present photometric, spectroscopic, and polarimetric observations of (163693) Atira collected on the night of 05 March 2024. A dual-band lightcurve acquired in the V and R_C filters using the Metsähovi Observatory (L08) in Kirkkonummi, Finland spans nearly 8~hours. From these measurements we estimate a V-R_C = 0.43 +/- 0.07 mag. A visible spectrum and linear polarisation measurement were obtained with the Nordic Optical Telescope in La Palma, Spain. We employ a semi-empirical model to recover the spectral flux that was lost due to a mis-aligned spectroscopic slit. The corrected model reflectance spectrum indicates a primitive Cgh taxonomic type with a distinct absorption feature from 0.60 to 0.85 um with a depth of ~7.4 %. Direct comparison with laboratory spectra of meteorites indicates close matches with CM carbonaceous chondrites. We detect a large degree of linear polarisation at a phase angle of 83 degrees which is consistent with (3200) Phaethon and (155140) 2005 UD, but lower than (162173) Ryugu and (152679) 1998 KU2. We use a CM carbonaceous chondrite density to estimate its bulk porosity of 51 +/- 31 % from previously estimated bulk density. The uncertainty in bulk density and bulk porosity is dominated by the volume of the primary. Future near-infrared spectroscopic and thermal infrared observations can be used to refine the composition and bulk density. Finally, we discuss implications for ground-based physical characterisation studies of the Atira population.

Orbital degrees of freedom play a pivotal role in shaping the physical properties of two-dimensional (2D) van der Waals magnetic systems, strongly influencing electron-phonon coupling and intermediate-state dynamics. In this study, we present a comprehensive temperature and thickness-dependent Raman investigation of the single crystals of 2D van der Waals antiferromagnetic series (NixFe1-x)2P2S6. Utilizing Raman spectroscopy, we explore the interplay between local orbital excitations and lattice vibrations, observing higher-order phonon modes extending up to the fourth order. We observe an anomalously high and weakly temperature-dependent intensity ratio of these higher-order modes relative to low-energy first-order phonons. Theoretical cluster calculations and Franck-Condon modelling reveal that these features originate from resonantly enhanced multiphonon scattering mediated by localized intermediate Ni 3d8 multiplet excitations. The local orbital occupancy of these intermediate states couples strongly to lattice coordinates, providing a selective enhancement mechanism for specific phonon channels. Notably, these higher-order features are absent in the end-member Fe2P2S6. This distinction reflects the charge-transfer character of Ni2P2S6, which places optical transitions in resonance with local multiplet states, contrasted with the Mott-Hubbard insulator regime in Fe-rich analogues. Our findings clarify the microscopic origin of high-frequency Raman scattering in transition metal thiophosphates and underscore the central role of local orbital-phonon interactions in 2D correlated magnets.

Spatially constrained semantic navigation requires robots to identify targets specified not only by semantic categories but also by relations to surrounding objects. In unknown environments, resolving such goals requires efficient exploration together with sufficient target and contextual evidence for reliable relation verification. Existing methods leave relation-aware verification and multi-robot collaboration largely disconnected: relational navigation is predominantly single-agent, while multi-robot systems seldom coordinate distributed observations for instance-specific relation verification. We propose CoRelNav, whose core is coupling task-conditioned multi-robot exploration with candidate-driven collaborative verification. A spatial-semantic field converts task constraints, scene nodes, and object features into exploration utility; as candidate information accumulates, robots are reallocated toward complementary evidence under team navigation costs, while instance-consistent observations are aggregated across topology nodes. This coupling reduces redundant search and enables relation hypotheses to be resolved from distributed partial evidence that independent exploration or isolated-view verification can leave ambiguous. Experiments in photorealistic simulation demonstrate consistent improvements over representative baselines, with ablations validating the proposed exploration and verification mechanisms. We further deploy the complete system on two physical mobile robots, demonstrating its applicability to real-world collaborative navigation.

An edge weighting of a graph induces a coloring of its vertices in which the color of a vertex is the total weight of the edges incident with it. Such an edge weighting is proper if adjacent vertices always receive distinct colors. Deciding whether a graph admits a proper weighting is known to be NP-complete for the weight set $\{0,1\}$, and also for $\{1,2\}$. In recent work (arXiv:2604.12363) we showed that both problems are FPT parameterized by the vertex cover number $k$, but it was open – to the best of our knowledge – whether either parameterized problem had a polynomial kernel. In this work, we show that both problems have polynomial kernels when parameterized by $k$. We also show that both problems are W[1]-hard parameterized by treedepth, answering another question from our earlier work. We then study the pre-weighted versions of the two problems, in which the weights of some edges are fixed in advance, and the task is to extend the assignment to a proper weighting of the whole graph. We show that both pre-weighted problems are FPT parameterized by the vertex cover number $k$. For the $\{1,2\}$ version the running time is $2^{O(k \log k)} \cdot n$; for the $\{0,1\}$ version we obtain the same running time when every pre-weight is $1$, and a slower FPT algorithm in the general case. We also show that both pre-weighted problems are W[1]-hard parameterized by either of (i) the feedback vertex set number or (ii) the treedepth of the input graph. Since a graph with no pre-assigned weights is a special case, our algorithms for the pre-weighted versions solve the two original problems as well, in time $2^{O(k \log k)} \cdot n$, significantly improving on the bound of $2^{O(k^4)} \cdot n^{O(1)}$ from our earlier work.

Seven Hamamatsu R8520-506 photomultiplier tubes were characterized at room temperature and in liquid argon to evaluate their response and operating stability for liquid-argon light readout. A 405~nm pulsed light source was used to measure gain, peak-to-valley ratio, single-photoelectron charge resolution, and afterpulsing, while dark-count data were acquired with random triggers in the absence of active illumination. Gain and dark-count stability were monitored during a ten-day operating period in liquid argon. At a motherboard input voltage of 900~V, the extracted gains in liquid argon were approximately 27.8%–51.3% higher than the corresponding room-temperature values. For all seven tubes, the peak-to-valley ratios were higher and the single-photoelectron charge resolution improved in liquid argon. The dark count rates ranged from approximately 143 to 252~Hz at room temperature and from 166 to 199~Hz in liquid argon, with three tubes showing increases and four showing decreases. The mean number of detected afterpulses per selected primary event was approximately 21.0%–63.6% lower in liquid argon. The relative standard deviations across the ten daily means of gain and dark count rate did not exceed approximately 2.1% and 3.3%, respectively. The measurements provide a quantitative basis for considering R8520-506 tubes as photosensor candidates for liquid-argon detectors.

Human-written agent skills encode rich workflows for real-world problem solving, but are typically used as external inference-time instructions rather than internalized as reusable model capabilities. We introduce SkillGym, a framework that transforms these skills into executable, verifiable training environments for large language model agents. Its skill-to-task pipeline instantiates concrete tasks, verifies outcomes with code-based checkers, and assesses empirical skill dependence through contrastive executions. We construct and release 2,756 environments across 12 categories and collect 8,364 successful trajectories from multiple models and harnesses, averaging 49 tool calls and over 60k logged text tokens. These resources support supervised fine-tuning on verified workflows and reinforcement learning with outcome-based rewards. Under Claude Code, supervised fine-tuning improves Qwen3.5-35B-A3B by 199 Elo on GDPval-AA v2, 19.10 percentage points on Terminal-Bench 2.1, and 28.13 and 12.38 points on SkillsBench v1.1 with and without skills, respectively. Our 35B SkillGym-Agent reaches 51.47% on skill-assisted SkillsBench, exceeding reported scores for Claude Sonnet 4.6, GPT-5.4 Mini, and DeepSeek V4 Pro. Without skills, it also surpasses skill-assisted bases under Codex and Claude Code, suggesting reusable procedural competence.