Science & Mathematics

Positron Emission Tomography (PET) provides important metabolic information for tumor detection and staging but remains significantly less accessible than Computed Tomography (CT) due to its high cost, infrastructure requirements, and reliance on injected radioactive tracers. CT-to-PET image translation using generative models aims to address this disparity, yet state-of-the-art diffusion-based methods are limited by the scarcity of paired training data. To overcome this, we propose a domain-adapted latent diffusion framework that transfers natural-image priors using a relatively small dataset of 21,930 paired CT/PET slices, incorporating a signed power transform to resolve extreme intensity sparsity in PET data. Compared to CPDM, a state-ofthe-art baseline for CT-to-PET image translation, our approach yields superior performance in pixel-wise reconstruction and localized tumor fidelity, demonstrating effective transfer of natural-image priors to the medical domain. We further provide a systematic analysis of evaluation metric reliability for sparse PET data, showing that standard full-image metrics can be misleading indicators of translated image quality, and propose background-filtered and mask-restricted metrics as more clinically meaningful alternatives.

Let $d_{K_{s,t}}(n)$ be the largest fraction of all graphs on $[n]$ in a family containing no two distinct members whose symmetric difference is a copy of $K_{s,t}$. We determine this maximum density for every fixed pair of positive integers $s,t$. If both part sizes are odd, it is known that the density is exactly $1/2$. Otherwise, let $2h$ be the largest even part size. We prove $$ d_{K_{s,t}}(n)=Θ_{s,t}(n^{-h}). $$ For $h>1$, this improves the general $O_{s,t}(1/n)$ upper bound obtained from Alon's two-copy gluing theorem~\cite{Alon2024}. The sharp bounds for even stars were already proved by Alon and are recovered as special cases. The upper bound uses the Frankl--Füredi forbidden-intersection theorem. For the lower bound, we construct binary linear codes using BCH codes. When exactly one part size is even, we assign to each edge the sum of the binary vectors assigned to its endpoints. When both part sizes are even, we identify these vectors with elements of a finite field and assign to each edge the product of its endpoint elements. In both cases, every copy of $K_{s,t}$ has a nonzero syndrome.

Evaluating human-AI co-design requires testing whether coding agents can translate real users' multimodal requests into working application revisions. We introduce Co-Design Bench, a reference-free benchmark of 273 tasks curated from 37,463 deduplicated queries across 26,314 real user coding sessions. Its two tracks, Multimodal Editing (ME) and Repository Editing (RE), cover localized edits with optional image, audio, video, or document inputs and coordinated repository revisions. Tasks span three capability dimensions: Grounded Understanding, UI Transformation, and Functional Implementation. The tasks pair normalized requests and runnable pre-edit applications with 2,331 hidden expert-authored acceptance criteria. Our judge-then-refute protocol grades criteria using source and browser evidence, challenges positive verdicts with a separate LLM verifier, and audits visual and interaction claims without gold code or target screenshots. Under a shared agentic harness, the strongest of ten model endpoints passes an average of 69.9% of criteria per task but fully satisfies only 26.4% of tasks. All ten endpoints achieve lower criterion completion on repository edits spanning three or four change types than on those spanning one or two. Co-Design Bench provides a testbed for studying the gap between partial requirement satisfaction and complete task success under agentic execution for real-user application revisions.

Pinching-antenna systems (PASSs) are an emerging flexible antenna technology that enables the creation of line-of-sight (LoS) links by repositioning a radiating element along a dielectric waveguide (WG). Most existing studies of PASSs assume a purely LoS environment, despite the fact that many practical wireless environments exhibit scattering and spatially correlated shadow fading. In this paper, we present the first statistical analysis of PASS performance under Ricean fading with spatially correlated large-scale fading. Focusing on a fundamental single-user uplink PASS, we characterize the distribution and mean of the maximum achievable signal-to-noise ratio (SNR). For the classical exponentially correlated shadow fading scenario, we express the channel power process in terms of Brownian motion and derive highly accurate approximations for the maximum SNR distribution and mean. For a general class of smoothly correlated shadow fading processes, we derive an accurate expression for the SNR distribution by modeling the process in terms of the fade sojourn duration. To address the challenges of antenna position selection in fading environments, we propose radio environment map (REM)-based positioning strategies, including opportunistic random-look and grid-based constructions. Analytical results are validated by simulation, and the impacts of key system parameters, such as the Ricean $K$-factor, spatial correlation, WG length, and carrier frequency are investigated. The results provide insights into PASS operation beyond idealized LoS conditions and establish a general analytical framework for performance evaluation and antenna positioning in scattered environments.

Large language models (LLMs) often produce plausible but incorrect answers, motivating reinforcement learning methods that reward correct answers, penalize incorrect answers, and allow abstention. However, fixed outcome-based rewards do not distinguish incorrect responses with different confidence levels or account for differences in answering reliability across queries. Consequently, such rewards may inadequately penalize overconfident errors and fail to discourage unnecessary abstention. We propose CalibraRL, a reinforcement learning framework that incorporates response confidence and query-level answering evidence into reward design. For non-abstaining responses, CalibraRL combines correctness-based utility with a Brier-based penalty that encourages agreement between confidence and correctness. For abstentions, it estimates conditional answering reliability using a Beta posterior derived from the correctness of non-abstaining responses within each sampled group. The abstention reward increases with the posterior probability that the model's accuracy on non-abstaining responses is below a predefined threshold. CalibraRL reuses training rollouts without requiring additional annotations indicating whether the model knows the answer to each question. Experiments on three QA datasets demonstrate that CalibraRL improves answer precision and reduces hallucinations compared with strong baselines. Further analysis shows that CalibraRL reduces mean expected calibration error (ECE) by 30.8\% across two backbone models and three datasets, highlighting its potential to improve the reliability of retrieval-augmented generation (RAG) in web search.

Understanding the envelope of shape variation encompassing both sexual dimorphism and growth-related changes in living tissue is a non-trivial task. In this work, we use statistical shape models (SSMs) and analyze demographic-stratified shape parameters to quantify sexual dimorphism and sex-specific shape changes. To this end, we developed a two-stage detection-segmentation framework that accurately segmented $n = 27664$ unique bone masks from a radiographic dataset of $n = 1456$ subjects with diverse ages and sexes (HD95 = $1.512 \pm 0.005$, ASSD $0.449 \pm 001$, Dice = $0.987 \pm 0.00004$, IoU = $0.975 \pm 0.0001$). Following SSM fitting, our follow-up analyses of stratified shape parameters highlighted the morphology of sexual dimorphisms for each bone and demonstrated sex-specific differences in shape development. Our results were consistent with existing literature and provided an unambiguous, data-driven methodology for characterizing tissue shape and its population-scale variation. This approach provides a potential methodological foundation for developing population-based models of anatomical variation and could be extended to investigate tissue growth and shape changes over time. Future work should evaluate the applicability of this methodology on alternative anatomical structures and imaging modalities, with the goal of validating these growth trajectories on longitudinal data.

Hardware-aware training often optimizes FPGA resources, latency, or bit operations, but folding changes what compression saves: shared arithmetic decouples precision, weight storage, and scheduled work. We derive training objectives from the measured costs of a folded streaming speech-enhancement accelerator. Mixed-precision weights reduce on-chip storage; structured channel sparsity removes entire operations from the engine schedule. We deploy a bit-exact implementation on the Kria KV260. The best sparse model eliminates 47.3% of convolution MACs, cutting frame time by 45% (1.86 to 1.03 ms) and dynamic energy by 60% (83 to 33 $μ$J/frame) against a storage-priced model measured in the same session, with higher speech quality. Relative to the most efficient dense build measured separately, frame time falls by 42% and energy by 48%. Reducing weight width from 8 to 6 bits lowers dynamic power by 27% but changes frame time by less than 5%. Plain channel group-lasso removes more work than cycle-cost-weighted lasso. Thus, folded-accelerator training should price storage and removable work separately, rather than rely on a single bit-operation metric.

AI data center build-out is a multi-trillion-dollar, multi-decade infrastructure program. Yet siting, power procurement, and cooling decisions are still made with static, single-point scoring tools. These tools are evaluated once, on day-zero numbers. This is a poor fit. A site's water, power, and carbon profile drifts substantially over the 10 to 20 year asset life. The governing decisions are also irreversible, and they bind capital and physical plant for years. We formalize long-horizon build-out as a tiered multi-objective Markov decision process. Strategic commitments are gated to discrete windows, incur switching costs, and energize only after source-dependent lead times. Operational dispatch is restricted to capacity that physically exists. We solve the problem with a preference-conditioned multi-objective agent that has a semi-Markov manager and worker hierarchy. The agent recovers a family of policies that a planner can steer across cost, carbon, water, and reliability at inference time. We validate long-horizon dynamics with a lightweight counterfactual rollout model. Across three commitment regimes (irreversible, reversible, and commit-once) on real site data, preference conditioning yields a confirmed, steerable trade-off on both objective pairs in every regime. Imposing lock-in produces a more evenly covered frontier and a far more consistent rate of anticipatory retrofit ahead of a binding water-stress constraint than a freely reversible formulation. Finally, a compact counterfactual dynamics model tracks slow-moving state well. It is measurably outperformed by simple persistence on the water-stress axis. We report this limitation rather than obscure it. We release the environment, agents, baselines, and evaluation harness. All of them are grounded in real interconnection-queue, water-stress, and grid-emissions data.

Recent advances in generative video models have improved visual realism, making detection increasingly dependent on subtle temporal inconsistencies. However, these signals are weak and spatially localized, while most of the video content is redundant. This creates a fundamental mismatch: existing detectors rely on heavy backbones to model full videos, which leads to high computational cost and limited scalability. In this work, we rethink AI-generated video detection from a representation perspective. Instead of modeling entire videos, we restructure them into compact temporal slice representations that isolate informative temporal evolution. We introduce a lightweight framework using frequency-aware temporal slice consistency learning, where informative spatial rows and columns are aggregated over time to form slice-domain inputs. This reparameterization suppresses redundant appearance while preserving discriminative temporal dynamics, enabling efficient and targeted detection. By constructing slices from both high- and low-frequency regions, our method exposes complementary temporal artifacts, including unstable details and abnormal motion smoothness. We instantiate this idea in a Temporal Slice Consistency Network, which integrates task-driven slice localization, hierarchical positional encoding, and a lightweight Transformer to model cross-slice and cross-frequency dependencies. We further introduce AI-Artist, a new benchmark with artist-curated videos from recent high-fidelity generators, including Seedance and Kling. Experiments on GenVideo and AI-Artist show that our method achieves strong performance while requiring over 400$\times$ fewer model-forward FLOPs and 3.6$\times$ faster model inference than the strong video-based baseline ReStraV (2.9$\times$ faster end to end, including preprocessing), enabling scalable and practical deployment.

Long-horizon web search requires LLM-based agents to integrate information across web pages through extended interactions under task-specific constraints. Evaluating this capability is important not only for measuring final answer quality, but also for assessing whether an agent acquires, verifies, and uses evidence reliably over extended interactions. Existing evaluations, however, mainly rely on simulated trajectories or on post-hoc trajectories reconstructed from final answers, rather than on retained real interaction traces, which limits the fidelity of behavioral assessment and weakens support for fine-grained error localization. To address this gap, we introduce LONGSEARCHEVAL, a trajectory-aware agent evaluation framework that combines a unified agent execution protocol with source constrained outcome verification and retrospective failure-node localization. Stage 1 verifies answer groups against agent-cited webpages without human-authored gold answers; Stage 2 audits normalized rounds and links flagged failure nodes to incorrect answer groups. We evaluate 16 models from eight families on 100 tasks, yielding 1,600 retained runs and 6,545 automatically flagged failure nodes. Trajectory collection and evaluation required 1,037 and 318 hours of cumulative runtime, respectively. Human validation yields 94.50% agreement on supplied answer-group labels and 98.11% feature-correspondence precision for audited round-group links. Together, outcome scores and linked trajectory evidence support controlled model comparison and retrospective inspection of answer-related content.

Host-based intrusion detectors order events, window audit streams and sequence attack stages by their timestamps. Every conclusion they draw therefore rests on assumptions about clocks, delivery, entity identity and the integrity of the evidence, and most of these assumptions go unstated. This Systematization of Knowledge (SoK) makes them explicit. Drawing on temporal databases, stream processing and runtime verification, we distill five components of a time model, together with a checklist of evidence assumptions and four levels of guarantee, and walk through a running campaign with them. A documented search yields fifteen families of detection and investigation. For papers selected by screening role and period, plus named systems and landmarks, we check the stated assumptions in the full texts, and we assess eleven temporal knowledge representation and reasoning (KRR) formalisms and a learned baseline on nine attack patterns. Some early alert correlators stated clock and delivery assumptions, and forensics measures clock error. Current provenance-based detectors state no clock bound. Four audited papers describe input-ordering conditions or conventions, but none specifies conditions for conclusion validity across a stated class of delivery failures. In the selected provenance-based systems, campaign composition uses system-specific procedures or objectives, learned state, causal graph queries or language models, and beyond CEP-based query systems, we found no temporal rule language or stream reasoner from KRR applied to host audit data. In 28 of 38 audited provenance-era detector papers, capture or stored evidence is trusted or its manipulation is placed out of scope, and none says which conclusions survive deleted, delayed or forged records. Five open challenges follow from these gaps.

Recent studies on generalizable human Gaussian splatting have achieved the notable progress in synthesizing novel views of unseen human subjects from multi-view images. Despite this progress, existing methods still struggle to represent fine-grained details of human subjects, which inherently involve high geometric and textural complexity. This is because previous methods do not impose a constraint on how densely Gaussians are placed, which often leads to blurry rendering results in complex regions. To address this problem, we propose a complexity-guided regularization scheme. The key idea is to leverage a complexity prior computed via the local entropy in geometric and textural characteristics of a clothed human. This effectively prevents Gaussians from being coarsely placed in complex regions, which facilitates the faithful reconstruction of intricate structures. Furthermore, we sample local patches predominantly from complex regions according to this complexity prior, which allows the model to be sufficiently supervised by local details. Experimental results on benchmark datasets demonstrate that our method delivers state-of-the-art performance while maintaining the real-time inference speed in generalizable human Gaussian splatting. The code and model are publicly available at https://github.com/DCVL-3D/CGR-GS_release.

Mn$_2$Au is one of the most promising materials for antiferromagnetic spintronics due to its high Néel temperature, moderate anisotropy, layered spin structure allowing current induced spin-orbit-torques, and ultrafast spin dynamics. Further understanding and control of the magnetic order in Mn$_2$Au for ultrafast spintronics applications via ultrashort electric/optical means requires accurate understanding of how its magnetic properties depend on temperature. Here we present atomistic spin dynamics simulations of the temperature-dependent macroscopic magnetic properties of Mn$_2$Au single crystals. We introduce a spin Hamiltonian based on previous ab initio calculations which produces temperature-dependent scaling of the anisotropy constants with magnetisation consistent with Zener-Akulov-Callen-Callen power laws. The calculated temperature dependence of exchange stiffness follows a scaling law with magnetisation of $m^{1.83}_e$ , and predicts a temperature-dependent domain wall width that agrees very well with direct simulation. Our model provides an understanding of temperature-dependent magnetic properties of Mn$_2$Au that enables large-scale simulation of the working conditions of spintronic devices.

The latest developments regarding the feasibility of muon-hadron collisions, achieved through the major reuse of LH$e$C infrastructure, are presented - at the LH$μ$C, a 1 or 2 TeV $μ^+$ beam collides with a 7 TeV proton beam, resulting in the $μ^+p$ centre-of-mass energy of 5.3 or 7.5 TeV, respectively, at an average luminosity close to $10^{34}$cm$^{-2}$s$^{-1}$. First, a concise scientific motivation is given, including exceptional prospects for neutrino science, and is followed by a presentation of the LH$μ$C concept in a nutshell - from the source of ultra-cold anti-muons, through the muon booster, to the main muon ring. Basic considerations regarding the machine-detector interface and the general-purpose detector are also presented. The paper concludes with an outlook underlying unique and powerful synergies between the LH$μ$C, LH$e$C, FCC and Muon Collider projects.

Inventories of open science initiatives differ in scope, geographic coverage and classification, making it difficult to compare infrastructures, services, networks and policies. We developed COSMI Hub within the OPENIT project to bring these sources together in a curated dataset. The release archived on 8 October 2026 contains 1,186 records and 71 variables covering initiative identity, source provenance, geography, record level, operational status, thematic activities, governance, funding and external identifiers. We identified candidates through structured directories, policy documents, official websites and suggestions from the open science community, then checked identities, reconciled duplicates and assessed supporting evidence. The dataset distinguishes standalone initiatives, services, deployments, umbrella organisations and policy instruments. Records can be assigned several thematic dimensions, with uncertainty retained where evidence is incomplete. We describe the collection and coding methods, the data structure and file-level checks, together with considerations for reuse. The Zenodo release is fixed, whereas the companion dashboard is updated separately. Because we assembled the collection purposively, it cannot be used to estimate the worldwide number of open science initiatives.