Search

Mechanistic interpretability aims to understand the computational mechanisms underlying neural networks' capabilities in order to accomplish concrete scientific and engineering goals. Progress in this field thus promises to provide greater assurance over AI system behavior and shed light on exciting scientific questions about the nature of intelligence. Despite recent progress toward these goals, there are many open problems in the field that require solutions before many scientific and practical benefits can be realized: Our methods require both conceptual and practical improvements to reveal deeper insights; we must figure out how best to apply our methods in pursuit of specific goals; and the field must grapple with socio-technical challenges that influence and are influenced by our work. This forward-facing review discusses the current frontier of mechanistic interpretability and the open problems that the field may benefit from prioritizing.

Benchmarks are important tools for tracking the rapid advancements in large language model (LLM) capabilities. However, benchmarks are not keeping pace in difficulty: LLMs now achieve over 90\% accuracy on popular benchmarks like MMLU, limiting informed measurement of state-of-the-art LLM capabilities. In response, we introduce Humanity's Last Exam (HLE), a multi-modal benchmark at the frontier of human knowledge, designed to be the final closed-ended academic benchmark of its kind with broad subject coverage. HLE consists of 3,000 questions across dozens of subjects, including mathematics, humanities, and the natural sciences. HLE is developed globally by subject-matter experts and consists of multiple-choice and short-answer questions suitable for automated grading. Each question has a known solution that is unambiguous and easily verifiable, but cannot be quickly answered via internet retrieval. State-of-the-art LLMs demonstrate low accuracy and calibration on HLE, highlighting a significant gap between current LLM capabilities and the expert human frontier on closed-ended academic questions. To inform research and policymaking upon a clear understanding of model capabilities, we publicly release HLE at https://lastexam.ai., Comment: 25 pages, 6 figures

We present measurements of large-scale cosmic microwave background (CMB) E-mode polarization from the Cosmology Large Angular Scale Surveyor (CLASS) 90 GHz data. Using 115 det-yr of observations collected through 2024 with a variable-delay polarization modulator, we achieved a polarization sensitivity of $78\,\mathrm{\mu K\,arcmin}$, comparable to Planck at similar frequencies (100 and 143 GHz). The analysis demonstrates effective mitigation of systematic errors and addresses challenges to large-angular-scale power recovery posed by time-domain filtering in maximum-likelihood map-making. A novel implementation of the pixel-space transfer matrix is introduced, which enables efficient filtering simulations and bias correction in the power spectrum using the quadratic cross-spectrum estimator. Overall, we achieved an unbiased time-domain filtering correction to recover the largest angular scale polarization, with the only power deficit, arising from map-making non-linearity, being characterized as less than $3\%$. Through cross-correlation with Planck, we detected the cosmic reionization at $99.4\%$ significance and measured the reionization optical depth $\tau=0.053^{+0.018}_{-0.019}$, marking the first ground-based attempt at such a measurement. At intermediate angular scales ($\ell>30$), our results, both independently and in cross-correlation with Planck, remain fully consistent with Planck's measurements., Comment: 24 pages, 19 figures, 3 tables; submitted to ApJ

We confirm the planetary nature of a pair of transiting sub-Neptune exoplanets orbiting the bright F-type sub-giant star TOI-6054 ($V=8.02$, $K=6.673$) as a part of the OrCAS radial velocity survey using WIYN/NEID observations. We find that TOI-6054b and TOI-6054c have radii of $2.65 \pm 0.15$ $R_{\oplus}$ and $2.81 \pm 0.18$ $R_{\oplus}$, respectively, and masses of $12.4 \pm 1.7$ $M_{\oplus}$ and $9.2 \pm 2.0$ $M_{\oplus}$. The planets have zero-albedo equilibrium temperatures of $1360 \pm 33$ K and $1144 \pm 28$ K. The host star has expanded and will evolve off of the Main Sequence within the next $\sim$500 Myr, and the resulting increase in stellar luminosity has more than doubled the stellar flux the two planets receive compared to the start of the host star's main sequence phase. Consequently, TOI-6054b may be losing some of its primordial H/He atmosphere -- if it has one. Based on dynamical simulations performed using the orbital parameters of the two planets, TOI-6054b, and TOI-6054c are very likely in a 5:3 mean motion resonance. The TOI-6054 system thus has the potential to be an excellent candidate for future atmospheric follow-up observations, with two similarly sized sub-Neptunes around a bright star. We also estimate that if TOI-6054b is currently losing its H/He atmosphere this should be observable from space and from the ground., Comment: 23 pages, 11 figures, submitted to AJ, partially updated with referee's comments

The time-variable emission from the accretion flow of Sgr A*, the supermassive black hole at the Galactic Center, has long been examined in the radio-to-mm, near-infrared (NIR), and X-ray regimes of the electromagnetic spectrum. However, until now, sensitivity and angular resolution have been insufficient in the crucial mid-infrared (MIR) regime. The MIRI instrument on JWST has changed that, and we report the first MIR detection of Sgr A*. The detection was during a flare that lasted about 40 minutes, a duration similar to NIR and X-ray flares, and the source's spectral index steepened as the flare ended. The steepening suggests synchrotron cooling is an important process for Sgr A*'s variability and implies magnetic field strengths $\sim$40--70 Gauss in the emission zone. Observations at $1.3~\mathrm{mm}$ with the Submillimeter Array revealed a counterpart flare lagging the MIR flare by $\approx$10 minutes. The observations can be self-consistently explained as synchrotron radiation from a single population of gradually cooling high-energy electrons accelerated through (a combination of) magnetic reconnection and/or magnetized turbulence., Comment: Accepted for publication ApJL

Context propagation remains a central challenge in language model architectures, particularly in tasks requiring the retention of long-range dependencies. Conventional attention mechanisms, while effective in many applications, exhibit limitations in maintaining coherent contextual representations over extended sequences due to their reliance on discrete token interactions. A novel approach is introduced through the formulation of Intrinsic Tensor Field Propagation (ITFP), which models contextual relationships as continuous tensor fields distributed across token embeddings. The propagation dynamics are governed through differential equations that enable a structured flow of contextual information, augmenting the standard attention mechanism to enhance coherence and recall. A series of experiments conducted on an open-source transformer-based model demonstrate that ITFP provides measurable improvements in contextual retention, dependency resolution, and inference stability across various linguistic structures. Comparisons with baseline models reveal a reduction in syntactic inconsistencies and factual errors, while ablation studies indicate that the choice of propagation depth and integration strength significantly impacts model performance. Additional evaluations assessing domain generalization suggest that ITFP effectively adapts across different text genres, reinforcing its applicability beyond conventional language modeling tasks. Although computational trade-offs are introduced through the inclusion of tensor field computations, empirical findings suggest that the benefits in accuracy and coherence outweigh the increased processing demands.

Quantum simulations of many-body systems are among the most promising applications of quantum computers. In particular, models based on strongly-correlated fermions are central to our understanding of quantum chemistry and materials problems, and can lead to exotic, topological phases of matter. However, due to the non-local nature of fermions, such models are challenging to simulate with qubit devices. Here we realize a digital quantum simulation architecture for two-dimensional fermionic systems based on reconfigurable atom arrays. We utilize a fermion-to-qubit mapping based on Kitaev's model on a honeycomb lattice, in which fermionic statistics are encoded using long-range entangled states. We prepare these states efficiently using measurement and feedforward, realize subsequent fermionic evolution through Floquet engineering with tunable entangling gates interspersed with atom rearrangement, and improve results with built-in error detection. Leveraging this fermion description of the Kitaev spin model, we efficiently prepare topological states across its complex phase diagram and verify the non-Abelian spin liquid phase by evaluating an odd Chern number. We further explore this two-dimensional fermion system by realizing tunable dynamics and directly probing fermion exchange statistics. Finally, we simulate strong interactions and study dynamics of the Fermi-Hubbard model on a square lattice. These results pave the way for digital quantum simulations of complex fermionic systems for materials science, chemistry, and high-energy physics., Comment: 8 pages, 5 figures. Methods: 15 pages, 9 figures

Artificial intelligence has significantly advanced skin cancer diagnosis by enabling rapid and accurate detection of malignant lesions. In this domain, most publicly available image datasets consist of single, isolated skin lesions positioned at the center of the image. While these lesion-centric datasets have been fundamental for developing diagnostic algorithms, they lack the context of the surrounding skin, which is critical for improving lesion detection. The iToBoS dataset was created to address this challenge. It includes 16,954 images of skin regions from 100 participants, captured using 3D total body photography. Each image roughly corresponds to a $7 \times 9$ cm section of skin with all suspicious lesions annotated using bounding boxes. Additionally, the dataset provides metadata such as anatomical location, age group, and sun damage score for each image. This dataset aims to facilitate training and benchmarking of algorithms, with the goal of enabling early detection of skin cancer and deployment of this technology in non-clinical environments., Comment: Article Submitted to Scientific Data

Blockchain ecosystems -- such as those built around chains, layers, and services -- try to engage users for a variety of reasons: user education, growing and protecting their market share, climbing metric-measuring leaderboards with competing systems, demonstrating usage to investors, and identifying worthy recipients for newly created tokens (airdrops). A popular approach is offering user quests: small tasks that can be completed by a user, exposing them to a common task they might want to do in the future, and rewarding them for completion. In this paper, we capture blockchain data from one deployed quest system that offered 43 unique quests over 10 months with 80M completions. We use this data to offer insight about the factors that impact task completion: amount of reward, monetary value of reward, difficulty, and cost. We also discuss the role of farming and bots, and the factors that complicate distinguishing real users from automated scripts.

Simulating quantum systems with their environments often requires non-unitary operations, and mapping these to quantum devices often involves expensive dilations or prohibitive measurement costs to achieve desired precisions. Building on prior work with a finite-differences strategy, we introduce an efficient and exact single-ancilla unitary decomposition technique that addresses these challenges. Our approach is based on Lagrange-Sylvester interpolation, akin to analytical differentiation techniques for functional interpolation. As a result, we can exactly express any arbitrary non-unitary operator with no finite approximation error using an easily computable decomposition. This can lead to several orders of magnitude reduction in the measurement cost, which is highly desirable for practical quantum computations of open systems., Comment: 8 pages, 5 figures

Dependent pattern matching is a key feature in dependently typed programming. However, there is a theory-practice disconnect: while many proof assistants implement pattern matching as primitive, theoretical presentations give semantics to pattern matching by elaborating to eliminators. Though theoretically convenient, eliminators can be awkward and verbose, particularly for complex combinations of patterns. This work aims to bridge the theory-practice gap by presenting a direct categorical semantics for pattern matching, which does not elaborate to eliminators. This is achieved using sheaf theory to describe when sets of arrows (terms) can be amalgamated into a single arrow. We present a language with top-level dependent pattern matching, without specifying which sets of patterns are considered covering for a match. Then, we give a sufficient criterion for which pattern-sets admit a sound model: patterns should be in the canonical coverage for the category of contexts. Finally, we use sheaf-theoretic saturation conditions to devise some allowable sets of patterns. We are able to express and exceed the status quo, giving semantics for datatype constructors, nested patterns, absurd patterns, propositional equality, and dot patterns., Comment: To appear at ESOP 2025

Large language models (LLMs) contain substantial factual knowledge which is commonly elicited by multiple-choice question-answering prompts. Internally, such models process the prompt through multiple transformer layers, building varying representations of the problem within its hidden states. Ultimately, however, only the hidden state corresponding to the final layer and token position are used to predict the answer label. In this work, we propose instead to learn a small separate neural network predictor module on a collection of training questions, that take the hidden states from all the layers at the last temporal position as input and outputs predictions. In effect, such a framework disentangles the representational abilities of LLMs from their predictive abilities. On a collection of hard benchmarks, our method achieves considerable improvements in performance, sometimes comparable to supervised fine-tuning procedures, but at a fraction of the computational cost., Comment: Accepted at AISTATS 2025

We present ultraviolet (UV) spectroscopic observations covering three distinct accretion states of the low-mass X-ray binary (LMXB) MAXI J1820+070: the luminous hard state, a hard-intermediate state and the soft state. Our observations were obtained during the 2018 eruption of MAXI J1820+070 with the Hubble Space Telescope (HST) and AstroSat observatory. The extinction towards the source turns out to be low - $\rm E_{B-V} = 0.2 \pm 0.05$ - making it one of the best UV accretion laboratories among LMXBs. Remarkably, we observe only moderate differences between all three states, with all spectra displaying similar continuum shapes and emission lines. Moreover, the continua are not well-described by physically plausible irradiated disc models. All of this challenges the standard reprocessing picture for UV emission from erupting LMXBs. The UV emission lines are double-peaked, with high-ionization lines displaying higher peak-to-peak velocities. None of the lines display obvious outflow signatures, even though blue-shifted absorption features have been seen in optical and near-infrared lines during the hard state. The emission line ratios are consistent with normal abundances, suggesting that the donor mass at birth was low enough to avoid CNO processing ($\rm M_{2,i} \lesssim 1.0 - 1.5 {\mathrm M_{\odot}}$). Finally, we study the evolution of UV variability in our time-resolved HST observations (hard and hard-intermediate states). All UV power spectra can be modelled with a broken power-law, superposed on which we tentatively detect the $\simeq 18$s quasi-periodic oscillation (QPO) that has been seen in other spectral bands., Comment: 19 pages, 14 figures, submitted to MNRAS

Secure Multi-Party Computation (MPC) is an important enabling technology for data privacy in modern distributed applications. We develop a new type theory to automatically enforce correctness,confidentiality, and integrity properties of protocols written in the \emph{Prelude/Overture} language framework. Judgements in the type theory are predicated on SMT verifications in a theory of finite fields, which supports precise and efficient analysis. Our approach is automated, compositional, scalable, and generalizes to arbitrary prime fields for data and key sizes., Comment: arXiv admin note: text overlap with arXiv:2407.16504

One of the primary objectives of the SPHEREx mission is to understand the origin of molecules such as H2O, CO2, and other volatile compounds at the early stages of planetary system formation. Because the vast majority of these compounds -- typically exceeding 95% -- exist in the solid phase rather than the gaseous phase in the systems of concern here, the observing strategy planned to characterize them is slightly unusual. Specifically, SPHEREx will target highly obscured sources throughout the Milky Way, and observe the species of concern in absorption against background illumination. SPHEREx spectrophotometry will yield ice column density measurements for millions of obscured Milky Way sources of all ages and types. By correlating those column densities with source ages, the SPHEREx mission will shed light on whether those molecules were formed in situ along with their nascent stellar systems, or whether instead they formed elsewhere and were introduced into those systems after their formation. To that end, this work describes version 7$.$1 of the SPHEREx Target List of Ice Sources (SPLICES) for the community. It contains about 8$.$6 million objects brighter than W2~12 Vega mag over much of the sky, principally within a broad strip running the length of the Milky Way midplane, but also within high-latitude molecular clouds and even the Magellanic Clouds., Comment: Published by ApJ. 21 pages, 6 figures. This article documents the original version of SPLICES (7.1). The current version as well as the complete catalog is publicly available along with release notes documenting all additions and changes at the NASA/IPAC Infrared Science Archive (IRSA) at this URL: https://irsa.ipac.caltech.edu/data/SPHEREx/SPLICES/

Global, three-dimensional, magnetohydrodynamic simulations of Rayleigh-Taylor instabilities at the disk-magnetosphere boundary of rotating, magnetized, compact stellar objects reveal that accretion occurs in three regimes: the stable regime, the chaotic unstable regime, and the ordered unstable regime. Here we track stochastic fluctuations in the pulse period $P(t)$ and aperiodic X-ray luminosity $L(t)$ time series of 24 accretion-powered pulsars in the Small Magellanic Cloud using an unscented Kalman filter to analyze Rossi X-ray Timing Explorer data. We measure time-resolved histories of the magnetocentrifugal fastness parameter $\omega(t)$ and we connect $\omega(t)$ with the three Rayleigh-Taylor accretion regimes. The 24 objects separate into two distinct groups, with 10 accreting in the stable regime, and 14 accreting in the ordered unstable regime. None of the 24 objects except SXP 293 visit the chaotic unstable regime for sustained intervals, although several objects visit it sporadically. The Kalman filter output also reveals a positive temporal cross-correlation between $\omega(t)$ and the independently measured pulse amplitude $A(t)$, which agrees with simulation predictions regarding the pulse-forming behavior of magnetospheric funnel flows in the three accretion regimes., Comment: 41 pages, 25 figures, accepted for publication in The Astrophysical Journal

Quantum state tomography (QST) allows for the reconstruction of quantum states through measurements and some inference technique under the assumption of repeated state preparations. Bayesian inference provides a promising platform to achieve both efficient QST and accurate uncertainty quantification, yet is generally plagued by the computational limitations associated with long Markov chains. In this work, we present a novel Bayesian QST approach that leverages modern distributed parallel computer architectures to efficiently sample a $D$-dimensional Hilbert space. Using a parallelized preconditioned Crank--Nicholson Metropolis-Hastings algorithm, we demonstrate our approach on simulated data and experimental results from IBM Quantum systems up to four qubits, showing significant speedups through parallelization. Although highly unorthodox in pooling independent Markov chains, our method proves remarkably practical, with validation ex post facto via diagnostics like the intrachain autocorrelation time. We conclude by discussing scalability to higher-dimensional systems, offering a path toward efficient and accurate Bayesian characterization of large quantum systems., Comment: 8 pages, 5 figures

In this paper, we extend the classical SIRS (Susceptible-Infectious-Recovered-Susceptible) model from mathematical epidemiology by incorporating a vaccinated compartment, V, accounting for an imperfect vaccine with waning efficacy over time. The SIRSV-model divides the population into four compartments and introduces periodic re-vaccination for waning immunity. The efficiency of the vaccine is assumed to decay with the time passed since the vaccination. Periodic re-vaccinations are applied to the population. We develop a partial differential equation (PDE) model for the continuous vaccination time and a coupled ordinary differential equation (ODE) system when discretizing the vaccination period. We analyze the equilibria of the ODE model and investigate the linear stability of the disease-free equilibrium (DFE). Furthermore, we explore an optimization framework where vaccination rate, re-vaccination time, and non-pharmaceutical interventions (NPIs) are control variables to minimize infection levels. The optimization objective is defined using different norm-based measures of infected individuals. A numerical analysis of the model's dynamic behavior under varying control parameters is conducted using path-following methods. The analysis focuses on the impacts of vaccination strategies and contact limitation measures. Bifurcation analysis reveals complex behaviors, including bistability, fold bifurcations, forward and backward bifurcations, highlighting the need for combined vaccination and contact control strategies to manage disease spread effectively.

Random (mixed) unitary channels describe an important subset of quantum channels, which are commonly used in quantum information, noise modeling, and quantum error mitigation. Despite their usefulness, there is substantial complexity in characterizing or identifying generic random unitary channels. We present a procedure for learning a class of random unitary channels on orthogonal unitary bases on a quantum computer utilizing Pauli learning and a contracted quantum learning procedure. Our approach involves a multi-objective, Pauli- and unitary-based minimization, and allows for learning locally equivalent channels. We demonstrate our approach for varying degrees of noise and investigate the scalability of these approaches, particularly with sparse noise models., Comment: 14 pages, 6 figures

We report the results of a rapid follow-up campaign on the Type IIb Supernova (SN) 2022hnt. We present a daily, multi-band, photometric follow-up using the Las Cumbres Observatory, the Zwicky Transient Facility, the orbiting \textit{Swift} observatory, and the Asteroid Terrestrial-impact Last Alert System (ATLAS). A distinctive feature in the light curve of SN 2022hnt and other IIb SNe is an early narrow peak prior to the ${}^{56}$Ni peak caused by rapid shock cooling of the hydrogen envelope, which can serve as an important probe of the properties of the massive progenitor star in the moments before explosion. Using SN 2022hnt as a case study, we demonstrate a framework of considerations for the application of shock cooling models to type IIb SNe, outlining a consistent procedure for future surveys of Type IIb SNe progenitor and explosion properties. \hll{We fit several recent models of shock-cooling emission and obtain progenitor radii between $\sim50$ and $\sim100$ $R_\odot$, as well as hydrogen-enriched envelope masses between $\sim0.01$ and $\sim0.1$ $M_\odot$, both consistent with values for other IIb SNe. One of these models is the model of \cite{Morag2023}, marking the first time this model has been applied to a Type IIb SN.} We evaluate contrasting predictions between shock-cooling models to construct a fiducial parameter set which can be used for comparison to other SNe. Finally, we investigate the possibility of extended wind breakout or precursor emission captured in the earliest detections., Comment: 19 pages, 11 figures; accepted for publication in ApJ

This study explores a novel approach to advancing dementia care by integrating socially assistive robotics, reinforcement learning (RL), large language models (LLMs), and clinical domain expertise within a simulated environment. This integration addresses the critical challenge of limited experimental data in socially assistive robotics for dementia care, providing a dynamic simulation environment that realistically models interactions between persons living with dementia (PLWDs) and robotic caregivers. The proposed framework introduces a probabilistic model to represent the cognitive and emotional states of PLWDs, combined with an LLM-based behavior simulation to emulate their responses. We further develop and train an adaptive RL system enabling humanoid robots, such as Pepper, to deliver context-aware and personalized interactions and assistance based on PLWDs' cognitive and emotional states. The framework also generalizes to computer-based agents, highlighting its versatility. Results demonstrate that the RL system, enhanced by LLMs, effectively interprets and responds to the complex needs of PLWDs, providing tailored caregiving strategies. This research contributes to human-computer and human-robot interaction by offering a customizable AI-driven caregiving platform, advancing understanding of dementia-related challenges, and fostering collaborative innovation in assistive technologies. The proposed approach has the potential to enhance the independence and quality of life for PLWDs while alleviating caregiver burden, underscoring the transformative role of interaction-focused AI systems in dementia care., Comment: 18 pages, 12 figures

Although nodal spin-triplet topological superconductivity appears probable in UTe2, its superconductive order-parameter $\Delta_k$ remains unestablished. In theory, a distinctive identifier would be the existence of a superconductive topological surface band (TSB), which could facilitate zero-energy Andreev tunneling to an s-wave superconductor, and also distinguish a chiral from non-chiral $\Delta_k$ via enhanced s-wave proximity. Here we employ s-wave superconductive scan-tips and detect intense zero-energy Andreev conductance at the UTe2 (0-11) termination surface. Imaging reveals sub-gap quasiparticle scattering interference signatures with a-axis orientation. The observed zero-energy Andreev peak splitting with enhanced s-wave proximity, signifies that $\Delta_k$ of UTe2 is a non-chiral state: B1u, B2u or B3u. However, if the quasiparticle scattering along the a-axis is internodal, then a non-chiral B3u state is the most consistent for UTe2., Comment: 45 pages, 5 figures, to appear in Science (2025)

ArchNEMESIS is an open-source Python package developed for the analysis of remote sensing spectroscopic observations of planetary atmospheres. It is based on the widely used NEMESIS radiative transfer and retrieval tool, which has been extensively used for the investigation of a wide variety of planetary environments. The main goal of archNEMESIS is to provide the capabilities of its Fortran-based predecessor, keeping or exceeding the efficiency in the calculations, and benefitting from the advantages Python tools provide in terms of usability and portability. The code, stored in a public GitHub repository under a GPL-v3.0 license, is accompanied by detailed documentation available at https://archnemesis.readthedocs.io/., Comment: Submitted 27 January 2025

GaN has emerged to be a major semiconductor akin to silicon due to its revolutionary impacts in solid state lighting, critically enabled by p-type doping, and high-performance radio-frequency and power electronics. Suffering from inefficient hole doping and low hole mobility, quantum oscillations in p-type GaN have not been observed, hindering fundamental studies of valence bands and hole transport in GaN. Here, we present the first observation of quantum oscillations of holes in GaN. Shubnikov-de Haas (SdH) oscillations in hole resistivity are observed in a quantum-confined two-dimensional hole gas at a GaN/AlN interface, where polarization-induced doping overcomes thermal freeze-out, and a sharp and clean interface boosts the hole mobility enough to unmask the quantum oscillations. These holes degenerately occupy the light and heavy hole bands of GaN and have record-high mobilities of ~1900 cm2/Vs and ~400 cm2/Vs at 3K, respectively. We use magnetic fields up to 72 T to resolve SdH oscillations of holes from both valence bands to extract their respective sheet densities, quantum scattering times, and the effective masses of light holes (0.5-0.7 m0) and heavy holes (1.9 m0). SdH oscillations of heavy and light holes in GaN constitute a direct metrology of valence bands and open new venues for quantum engineering in this technologically important semiconductor. Like strained silicon transistors, strain-engineering of the valence bands of GaN is predicted to dramatically improve hole mobilities by reducing the hole effective mass, a proposal that can now be explored experimentally, particularly in a fully fabricated transistor, using quantum oscillations. Furthermore, the findings of this work suggest a blueprint to create 2D hole gases and observe quantum oscillations of holes in related wide bandgap semiconductors such as SiC and ZnO in which such techniques are not yet possible.

In distributed and federated learning, heterogeneity across data sources remains a major obstacle to effective model aggregation and convergence. We focus on feature heterogeneity and introduce energy distance as a sensitive measure for quantifying distributional discrepancies. While we show that energy distance is robust for detecting data distribution shifts, its direct use in large-scale systems can be prohibitively expensive. To address this, we develop Taylor approximations that preserve key theoretical quantitative properties while reducing computational overhead. Through simulation studies, we show how accurately capturing feature discrepancies boosts convergence in distributed learning. Finally, we propose a novel application of energy distance to assign penalty weights for aligning predictions across heterogeneous nodes, ultimately enhancing coordination in federated and distributed settings., Comment: 15 pages, 5 figures

The rise of TikTok as a primary source of information for youth, combined with its unique short-form video format, creates urgent questions about how substance use content manifests and spreads on the platform. This paper provides the first in-depth exploration of substance use-related content on TikTok, covering all major substance categories as classified by the Drug Enforcement Agency. Through social network analysis and qualitative coding, we examined more than 2,333 hashtags across 39,509 videos, identified 16 distinct hashtag communities and analyzed their interconnections and thematic content. Our analysis revealed a highly interconnected small-world network where recovery-focused hashtags like #addiction, #recovery, and #sober serve as central bridges between communities. Through manual coding of 351 representative videos, we found that Recovery Advocacy content (33.9%) and Satirical content (28.2%) dominate, while direct substance depiction appears in only 26% of videos, with active use shown in just 6.5% of them. This suggests TikTok functions primarily as a recovery support platform rather than a space promoting substance use. We found strong alignment between hashtag communities and video content, indicating organic community formation rather than attempts to evade content moderation. Our findings inform how platforms can balance content moderation with preserving valuable recovery support communities, while also providing insights for the design of social media-based recovery interventions.

We define what it means for a joint probability distribution to be compatible with a set of independent causal mechanisms, at a qualitative level -- or, more precisely, with a directed hypergraph ${\mathcal{A}}$, which is the qualitative structure of a probabilistic dependency graph (PDG). When ${\mathcal{A}}$ represents a qualitative Bayesian network, QIM-compatibility with ${\mathcal{A}}$ reduces to satisfying the appropriate conditional independencies. But giving semantics to hypergraphs using QIM-compatibility lets us do much more. For one thing, we can capture functional dependencies. For another, we can capture important aspects of causality using compatibility: we can use compatibility to understand cyclic causal graphs, and to demonstrate structural compatibility, we must essentially produce a causal model. Finally, QIM-compatibility has deep connections to information theory. Applying our notion to cyclic structures helps to clarify a longstanding conceptual issue in information theory., Comment: NeurIPS 2024

Background: When developing a clinical prediction model using time-to-event data, previous research focuses on the sample size to minimise overfitting and precisely estimate the overall risk. However, instability of individual-level risk estimates may still be large. Methods: We propose a decomposition of Fisher's information matrix to examine and calculate the sample size required for developing a model that aims for precise and fair risk estimates. We propose a six-step process which can be used before data collection or when an existing dataset is available. Steps (1) to (5) require researchers to specify the overall risk in the target population at a key time-point of interest; an assumed pragmatic 'core model' in the form of an exponential regression model; the (anticipated) joint distribution of core predictors included in that model; and the distribution of any censoring. Results: We derive closed-form solutions that decompose the variance of an individual's estimated event rate into Fisher's unit information matrix, predictor values and total sample size; this allows researchers to calculate and examine uncertainty distributions around individual risk estimates and misclassification probabilities for specified sample sizes. We provide an illustrative example in breast cancer and emphasise the importance of clinical context, including risk thresholds for decision making, and examine fairness concerns for pre- and post-menopausal women. Lastly, in two empirical evaluations, we provide reassurance that uncertainty interval widths based on our approach are close to using more flexible models. Conclusions: Our approach allows users to identify the (target) sample size required to develop a prediction model for time-to-event outcomes, via the pmstabilityss module. It aims to facilitate models with improved trust, reliability and fairness in individual-level predictions., Comment: arXiv admin note: text overlap with arXiv:2407.09293

Large language model (LLM) inference workload dominates a wide variety of modern AI applications, ranging from multi-turn conversation to document analysis. Balancing fairness and efficiency is critical for managing diverse client workloads with varying prefix patterns. Unfortunately, existing fair scheduling algorithms for LLM serving, such as Virtual Token Counter (VTC), fail to take prefix locality into consideration and thus suffer from poor performance. On the other hand, locality-aware scheduling algorithms in existing LLM serving frameworks tend to maximize the prefix cache hit rate without considering fair sharing among clients. This paper introduces the first locality-aware fair scheduling algorithm, Deficit Longest Prefix Match (DLPM), which can maintain a high degree of prefix locality with a fairness guarantee. We also introduce a novel algorithm, Double Deficit LPM (D$^2$LPM), extending DLPM for the distributed setup that can find a balance point among fairness, locality, and load-balancing. Our extensive evaluation demonstrates the superior performance of DLPM and D$^2$LPM in ensuring fairness while maintaining high throughput (up to 2.87$\times$ higher than VTC) and low per-client (up to 7.18$\times$ lower than state-of-the-art distributed LLM serving system) latency.

Emerging two-dimensional transition metal dichalcogenides (TMDCs) offer a promising platform for on-chip integrated photonics because of their unique optical and electronic properties. Their naturally passivated surfaces make them highly tolerant to lattice mismatch, enabling seamless heterogeneous integration by stacking different van der Waals materials, a crucial step in the development of advanced photonic devices. Here, we demonstrate the use of an inverse design wavelength demultiplexing waveguides for on-chip sorting and routing of distinct photoluminescence from the heterojunction formed by WS2 and WSe2 monolayers. The integrated nanophotonic chip splits and sorts excitonic emission into individual waveguides at room temperature. Our demonstration opens up new perspectives for integrating light sources in van der Waals materials with functional integrated photonics, offering a versatile platform for both fundamental research and practical applications.

Magnetic Resonance Imaging (MRI) is an essential diagnostic tool in clinical settings, but its utility is often hindered by noise artifacts introduced during the imaging process.Effective denoising is critical for enhancing image quality while preserving anatomical structures. However, traditional denoising methods, which often assume uniform noise distributions, struggle to handle the non-uniform noise commonly present in MRI images. In this paper, we introduce a novel approach leveraging a sparse mixture-of-experts framework for MRI image denoising. Each expert is a specialized denoising convolutional neural network fine-tuned to target specific noise characteristics associated with different image regions. Our method demonstrates superior performance over state-of-the-art denoising techniques on both synthetic and real-world brain MRI datasets. Furthermore, we show that it generalizes effectively to unseen datasets, highlighting its robustness and adaptability., Comment: Accepted to the WACV Workshop on Image Quality

We analyze 23 spectroscopically confirmed Type-2 quasars (QSOs) selected from the WISE 22$\rm \mu$m band in the SDSS Stripe 82 region, focusing on their multi-band photometry and spectral energy distributions (SEDs). These objects were selected to be IR-luminous ($\rm flux_{W4} > 5mJy$, i.e., $12.62 < W4 < 14.62 \rm\ AB \, magnitude$), optically faint ($r > 23$) or with red color ($r - W4 >8.38$). Gemini/GNIRS observations were conducted for all 24 candidates, and 18/24 were also observed with Keck/LRIS. The observations confirm 23 to be real Type-2 QSOs in the redshift range $0.88 - 2.99$ (12 are at $z>2$). We collect multi-band photometry and conduct SED fitting. The composite photometry probes the wavelength from 0.1$\rm \mu$m to 10$\rm \mu$m at the rest frame. The IR emission is dominated by dust torus implying an average torus luminosity for the sample of $L_{\rm torus} 10^{46.84} \rm erg/s$. The origin of the rest-UV/optical light is not definitive, but we present three possible scenarios: scattered light, stellar emission, and the reddened accretion disk. Assuming an obscured:unobscured ratio of approximately 1:1, our targets have $L_{\rm bol} = 10^{46.28} \rm erg \,s^{-1} - 10^{47.49} \rm erg \,s^{-1}$ and around SMBH masses $\rm 10^{8.18} M_{\odot} - 10^{9.39} M_{\odot}$, assuming they accreate at the Eddington limit. Compared to previous Type-2 AGN SEDs, our targets have a brighter dust torus and redder optical-IR color. By comparing the SED to the results from JWST `little red dots' (LRDs), we find that these IR-selected Type-2 QSOs have similar SED shapes to the LRDs. This pilot Type-2 QSO survey demonstrates that mid-IR selection is an efficient way to find luminous Type-2 QSOs at $z>2$. Finally, the composite photometry and Type-2 QSOs SED model generated by this sample provide a guide for finding more Type-2 QSOs at higher redshift.

Extreme coronal line emitters (ECLEs) are a rare class of galaxy that exhibit strong, high-ionization iron coronal emission lines in their spectra. In some cases, these lines are transient and may be the result of tidal disruption event (TDEs). To test this connection, we calculate the rate of variable ECLEs (vECLEs) at redshift $\sim0.3$. We search for ECLEs in the Baryon Oscillation Spectroscopic Survey (BOSS) LOWZ sample and discover two candidate ECLEs. Using follow-up spectra from the Dark Energy Spectroscopic Instrument and Gemini Multi-Object Spectrograph, and mid-infrared observations from the Wide-field Infrared Survey Explorer, we determine that one of these galaxies is a vECLE. Using this galaxy, we calculate the galaxy-normalized vECLE rate at redshift $\sim0.3$ to be $R_\mathrm{G}=1.6~^{+3.8}_{-1.4}\times10^{-6}~\mathrm{galaxy}^{-1}~\mathrm{yr}^{-1}$ and the mass-normalized rate to be $R_\mathrm{M}=7~^{+16}_{-6}\times10^{-18}~\mathrm{M_\odot^{-1}}~\mathrm{yr}^{-1}$. This is then converted to a volumetric rate of $R_\mathrm{V}=1.8~^{+4.5}_{-1.5}\times10^{-9}~\mathrm{Mpc}^{-3}~\mathrm{yr}^{-1}$. Formally, the LOWZ vECLE rates are $2-4$ times lower than the rates calculated from the Sloan Digital Sky Survey Legacy sample at redshift $\sim0.1$. However, given the large uncertainties on both measurements, they are consistent with each other at $1\sigma$. Both the galaxy-normalized and volumetric rates are one to two orders of magnitude lower than TDE rates from the literature, consistent with vECLEs being caused by $5-20$ per cent of all TDEs., Comment: 15 pages, 10 figures. Submitted to MNRAS. arXiv admin note: text overlap with arXiv:2402.16951

The organization of latent knowledge within large-scale models poses unique challenges when addressing overlapping representations and optimizing contextual accuracy. Conceptual redundancies embedded across layers often result in inefficiencies that affect both computational demands and task-specific outcomes. A framework was proposed to restructure these redundancies through advanced clustering techniques and dynamic thresholding, ensuring that critical semantic relationships are preserved while removing unnecessary overlaps. Evaluations revealed improved memory efficiency and faster inference times, alongside better alignment in latent knowledge clusters that enhanced interpretability. Improvements in error rates and adversarial robustness suggest that restructuring redundancies has broader implications for increasing model reliability across diverse applications. Comparative analyses highlighted reductions in resource consumption and notable gains in performance, particularly in translation and summarization tasks. Energy metrics demonstrated significant savings during training phases, further validating the practicality of the approach for real-world deployments. Representational fidelity was also enhanced, with latent space evaluations indicating better cluster alignment and higher semantic consistency. The methodology bridges a key gap in model optimization through directly addressing redundancies at the structural level. Its application opens avenues for scalable, efficient, and contextually aware systems that can adapt to complex, domain-specific tasks without compromising on performance.

We introduce the stochastic Zassenhaus expansions (SZEs), a class of ancilla-free quantum algorithms for Hamiltonian simulation. These algorithms map nested Zassenhaus formulas onto quantum gates and then employ randomized sampling to minimize circuit depths. Unlike Suzuki-Trotter product formulas, which grow exponentially long with approximation order, the nested commutator structures of SZEs enable high-order formulas for many systems of interest. For a 10-qubit transverse-field Ising model, we construct an 11th-order SZE with 42x fewer CNOTs than the standard 10th-order product formula. Further, we empirically demonstrate regimes where SZEs reduce simulation errors by many orders of magnitude compared to leading algorithms., Comment: 16 pages, 3 figures

Active learning in computer experiments aims at allocating resources in an intelligent manner based on the already observed data to satisfy certain objectives such as emulating or optimizing a computationally expensive function. There are two main ingredients for active learning: an initial experimental design, which helps to approximately learn the function, and a surrogate modeling technique, which provides a prediction of the output along with its uncertainty estimates. Space-filling designs are commonly used as initial design and Gaussian processes for surrogate modeling. This article aims at improving the active learning procedure by proposing a new type of initial design and a new correlation function for the Gaussian process. The ideas behind them are known in other fields such as in sensitivity analysis or in kernel theory, but they never seem to have been used for active learning in computer experiments. We show that they provide substantial improvement to the state-of-the-art methods for both emulation and optimization. We support our findings through theory and simulations, and a real experiment on the vapor-phase infiltration process., Comment: 28 pages, 9 figures

Particle discretizations of partial differential equations are advantageous for high-dimensional kinetic models in phase space due to their better scalability than continuum approaches with respect to dimension. Complex processes collectively referred to as \textit{particle noise} hamper long-time simulations with particle methods. One approach to address this problem is particle mesh adaptivity or remapping, known as \textit{particle resampling}. This paper introduces a resampling method that projects particles to and from a (finite element) function space. The method is simple; using standard sparse linear algebra and finite element techniques, it can adapt to almost any set of new particle locations and preserves all moments up to the order of polynomial represented exactly by the continuum function space. This work is motivated by the Vlasov-Maxwell-Landau model of magnetized plasmas with up to six dimensions, $3X$ in physical space and $3V$ in velocity space, and is developed in the context of a $1X$ + $1V$ Vlasov-Poisson model of Landau damping with logically regular particle and continuum phase space grids. The evaluation codes are publicly available, along with the data and reproducibility artifacts, and developed in the PETSc numerical library.

The KOTO II experiment is proposed to measure the branching ratio of the decay $K_L\to\pi^0\nu\bar{\nu}$ at J-PARC. With a beamline to extract long-lived neutral kaons at 5 degrees from a production target, the single event sensitivity of the decay is $8.5\times 10^{-13}$, which is much smaller than the Standard Model prediction $3\times 10^{-11}$. This allows searches for new physics beyond the Standard Model and the first discovery of the decay with a significance exceeding $5\sigma$. As the only experiment proposed in the world dedicated to rare kaon decays, KOTO II will be indispensable in the quest for a complete understanding of flavor dynamics in the quark sector. Moreover, by combining efforts from the kaon community worldwide, we plan to develop the KOTO II detector further and expand the physics reach of the experiment to include measurements of the branching ratio of the $K_L\to\pi^0\ell^+\ell^-$ decays, studies of other $K_L$ decays, and searches for dark photons, axions, and axion-like particles. KOTO II will therefore obtain a comprehensive understanding of $K_L$ decays, providing further constraints on new physics scenarios with existing $K^+$ results., Comment: Submitted to the J-PARC PAC. arXiv admin note: substantial text overlap with arXiv:2110.04462

Turbulent heat and momentum transfer processes due to thermal convection cover many scales and are of great importance for several natural and technical flows. One consequence is that a fully resolved three-dimensional analysis of these turbulent transfers at high Rayleigh numbers, which includes the boundary layers, is possible only using supercomputers. The visualization of these dynamics poses an additional hurdle since the thermal and viscous boundary layers in thermal convection fluctuate strongly. In order to track these fluctuations continuously, data must be tapped at high frequency for visualization, which is difficult to achieve using conventional methods. This paper makes two main contributions in this context. First, it discusses the simulations of turbulent Rayleigh-B\'enard convection up to Rayleigh numbers of $Ra=10^{12}$ computed with NekRS on GPUs. The largest simulation was run on 840 nodes with 3360 GPU on the JUWELS Booster supercomputer. Secondly, an in-situ workflow using ASCENT is presented, which was successfully used to visualize the high-frequency turbulent fluctuations., Comment: 16 pages, 15 figures, 6 tables

We investigate the pressure-driven structural and electronic evolution of PdSe\(_2\) using powder X-ray diffraction, Raman spectroscopy, and first-principles calculations. Beyond 2.3 GPa, suppression of the Jahn-Teller distortion induces in-plane lattice expansion and metallization. Around 4.8 GPa, the interlayer \(d_{z^2}-\pi^*\) orbital hybridization drives the dimensional crossover, facilitating the transformation from the 2D distorted to a 3D undistorted pyrite phase. Above 9 GPa, a novel phase emerges, characterized by octahedral distortions in the $d$ orbitals of Pd. Structural analysis suggests the presence of marcasite (\(Pnnm\)) or arsenopyrite (\(P2_1/c\)) phase with orthorhombic and monoclinic configurations, respectively. Furthermore, the observed phonon anomaly and electronic structure modifications, including the emergence of flat bands in the high-pressure phases, elucidate the fundamental mechanisms underlying the previously reported exotic superconductivity with an enhanced critical temperature. These results highlight the pivotal role of dimensional crossover and structural transitions in tuning the electronic properties of puckered materials, providing pathways for novel functionalities.

We study the piecewise constant bandit problem where the expected reward is a piecewise constant function with one change point (discontinuity) across the action space $[0,1]$ and the learner's aim is to locate the change point. Under the assumption of a fixed exploration budget, we provide the first non-asymptotic analysis of policies designed to locate abrupt changes in the mean reward function under bandit feedback. We study the problem under a large and small budget regime, and for both settings establish lower bounds on the error probability and provide algorithms with near matching upper bounds. Interestingly, our results show a separation in the complexity of the two regimes. We then propose a regime adaptive algorithm which is near optimal for both small and large budgets simultaneously. We complement our theoretical analysis with experimental results in simulated environments to support our findings., Comment: 44 pages, 7 figures

Seeking to unravel the enigmatic Verwey transition and its interplay with magnetism, we have conducted comprehensive measurements on the temperature-dependent electrical resistivity and magnetic moment of stoichiometric and doped-magnetite single crystals at temperatures reaching 1000 K. These investigations have allowed us to identify the Curie temperature, $T_C$, and other characteristic temperatures of the electrical resistivity. Remarkably, we have identified correlations between these temperatures and the Verwey temperature, $T_V$, indicating that the electrical transport properties and the mechanism of the Verwey transition are closely related to the magnetic properties.

We provide guidance on multiple imputation of missing at random treatments in observational studies. Specifically, analysts should account for both covariates and outcomes, i.e., not just use propensity scores, when imputing the missing treatments. To do so, we develop outcome-assisted multiple imputation of missing treatments: the analyst fits a regression for the outcome on the treatment indicator and covariates, which is used to sharpen the predictive probabilities for missing treatments under an estimated propensity score model. We derive an expression for the bias of the inverse probability weighted estimator for the average treatment effect under multiple imputation of missing treatments, and we show theoretically that this bias can be made small by using outcome-assisted multiple imputation. Simulations demonstrate empirically that outcome-assisted multiple imputation can offer better inferential properties than using the treatment assignment model alone. We illustrate the procedure in an analysis of data from the National Longitudinal Survey of Youth.

Given the hyperinflation that most of the Latin American countries suffered in the 90 and their decision towards adopting dollarization and in most cases keeping their own currency, this paper analyzes the effectiveness of dollarization as a protective mechanism against economic disruptions in Latin American countries. It assesses the context that led Latin American dollarized countries to dollarize and analyzes CPI, GDP, and the poverty rates pre, during, and postpandemic in Latin American countries, considering those that are dollarized and those that are not, and evaluating its relation to the US. Interviews were carried out with experts in the field. It assesses the advantages and disadvantages of dollarization regarding global crises. The data was compared and analyzed to check if there were patterns that support the paper objective which is that dollarization might serve as a protective mechanism against economic disruption. It was found that dollarization protects the economy against inflation, however, it does not fully protect the economy when considering economic performance and poverty. In conclusion, this research concludes that dollarization does not completely serve as a protective mechanism against economic disruptions nonetheless, it found that a bigger role is played by domestic policies and government action., Comment: 45 pages, 8 figures

Substance use disorders (SUDs) are a serious public health concern in the United States. Alcohol and cannabis are two of the most widely used substances. For adolescent/youth users of alcohol or cannabis, we propose a joint Bayesian learning model to predict their risks of developing alcohol use disorder (AUD) and cannabis use disorder (CUD) in adulthood based on their personal risk factors. The model is trained on nationally representative longitudinal data from Add Health (n = 12503). It consists of sub-models that predict the two SUDs for three groups of users-those who use alcohol only, cannabis only, and both substances - based on shared as well as unique risk factors. The model comprises of ten predictors. We externally validate the model on two independent datasets. The areas under the receiver operating characteristic curves for AUD and CUD, respectively, are: (a) 0.719 and 0.690 based on 5-fold cross-validation, (b) 0.748 and 0.710 based on validation dataset 1, and (c) 0.650 and 0.750 based on validation dataset 2. A simulation study shows that the proposed joint modeling approach generally performs better than separate univariate modeling of the corresponding dependent outcomes in terms of predictive accuracy. Our model may help in identifying adolescent substance users at high risk of developing SUD in adulthood, who can then be helped with appropriate intervention.

For many tasks, multi-robot teams often provide greater efficiency, robustness, and resiliency. However, multi-robot collaboration in real-world scenarios poses a number of major challenges, especially when dynamic robots must balance competing objectives like formation control and obstacle avoidance in the presence of stochastic dynamics and sensor uncertainty. In this paper, we propose a distributed, multi-agent receding-horizon feedback motion planning approach using Probably Approximately Correct Nonlinear Model Predictive Control (PAC-NMPC) that is able to reason about both model and measurement uncertainty to achieve robust multi-agent formation control while navigating cluttered obstacle fields and avoiding inter-robot collisions. Our approach relies not only on the underlying PAC-NMPC algorithm but also on a terminal cost-function derived from gyroscopic obstacle avoidance. Through numerical simulation, we show that our distributed approach performs on par with a centralized formulation, that it offers improved performance in the case of significant measurement noise, and that it can scale to more complex dynamical systems., Comment: 10 pages, 12 figures

Scenario-based testing is currently the dominant simulation-based validation approach for ADS. Its effective application raises two interrelated issues. The first is the choice of the method used to generate scenarios, based on various criteria such as risk, degree of autonomy, degree of coverage and representativeness, and complexity. The other is the choice of the evaluation method for estimating the safety and performance of the system under test. This work extends a study of the critical configuration testing (CCTest) approach we have already applied to four open modular autopilots. This approach differs from general scenario-based approaches in that it uses only realistic, potentially safe critical scenarios. It enables an accurate assessment of the ability to drive safely in critical situations for which feasible safety policies exist. Any incident observed in the simulation involves the failure of a tested autopilot. The contribution of this paper is twofold. First, we apply the critical configuration testing approach to four end-to-end open autopilots, Transfuser, InterFuser, MILE and LMDriver, and compare their test results with those of the four modular open autopilots previously tested with the same approach implemented in the Carla simulation environment. This comparison identifies both differences and similarities in the failures of the two autopilot types in critical situations. Secondly, we compare the evaluations of the four autopilots carried out in the Carla Leaderboard with our results obtained by testing critical configurations. This comparison reveals significant discrepancies, reflecting differences in test case generation criteria and risk assessment methods. It underlines the need to work towards the development of objective assessment methods combining qualitative and quantitative criteria.

We explore the evolving relationship between galaxies and their dark matter halos from $z \sim 0.1$ to $z \sim 12$ using mass-limited angular clustering measurements in the 0.54 deg$^2$ of the COSMOS-Web survey. This study provides the first measurements of the mass-limited two-point correlation function at $z \ge 10$ and a consistent analysis spanning 13.4 Gyr of cosmic history, setting new benchmarks for future simulations and models. Using a halo occupation distribution (HOD) framework, we derive characteristic halo masses and the stellar-to-halo mass relationship (SHMR) across redshifts and stellar mass bins. Our results first indicate that HOD models fit data at $z \ge 2.5$ best when incorporating a non-linear scale-dependent halo bias, boosting clustering at non-linear scales (r = 10-100 kpc). We find that galaxies at z > 10.5 with $\log(M_\star / M_\odot) \ge 8.85$ are hosted by halos with $M_{\rm h} \sim 10^{10.5}\,M_\odot$, achieving a star formation efficiency (SFE) $M_\star / (f_b M_{\rm h}) $ up to 1 dex higher than at $z \le 1$. The high galaxy bias at $z \ge 8$ suggests that these galaxies reside in massive halos with intrinsic high SFE. Our SHMR evolves significantly with redshift, starting high at $z \ge 10.5$, decreasing until $z \sim 2 - 3$, then increasing again until the present. Current simulations fail to reproduce both massive high-$z$ galaxies and this evolution, while semi-empirical models linking SFE to halo mass, accretion rates, and redshift align with our findings. We propose that $z > 8$ galaxies experience bursty star formation without significant feedback altering their growth, driving the rapid growth of massive galaxies observed by JWST. Over time, increasing feedback efficiency and exponential halo growth suppress star formation. At $z \sim 2 - 3$ and after, halo growth slows down while star formation continues, supported by gas reservoirs in halos., Comment: Submitted to Astronomy & Astrophysics

We investigated the temporal distribution of goals in soccer using event-level data from 3,433 matches across 21 leagues and competitions. Contrary to the prevailing notion of randomness, we found that the probability of a goal being scored is higher as matches progress, and we observed fewer-than-expected goals in the early minutes of each half. Further analysis of the time between subsequent goals shows an exponential decay, indicating that most goals naturally cluster closer together in time. By splitting this distribution by the team that scores the next goal, we observe bursty goal-scoring dynamics, wherein the same team is more likely to score again shortly after its previous goal. These findings highlight the importance of match context (whether driven by fatigue, tactical adaptations, or psychological momentum) in shaping when teams are able to score. Moreover, the results open avenues for extending data-driven methods for identifying high-impact moments in a match and refining strategic decision-making in soccer's evolving analytical landscape.