Search

Non-ergodicity impacts statistical inference in a diverse range of disciplines inside and outside of physics. However the concept of ergodicity is used inconsistently, and may refer to several nonequivalent notions. To help address this, we first identify and clarify the relationship between three major interpretations of ergodicity. We then introduce a method of spectral analysis of non-ergodicity which may be performed using data alone, and so may be applied in both numerical and empirical contexts. This may be used to identify, quantify, and classify non-ergodic populations within an ergodic decomposition. This is demonstrated with an application to the Kob-Andersen kinetically constrained lattice glass model., Comment: 6 pages, 3 figures

Recent advancements have positioned AI, and particularly Large Language Models (LLMs), as transformative tools for scientific research, capable of addressing complex tasks that require reasoning, problem-solving, and decision-making. Their exceptional capabilities suggest their potential as scientific research assistants but also highlight the need for holistic, rigorous, and domain-specific evaluation to assess effectiveness in real-world scientific applications. This paper describes a multifaceted methodology for Evaluating AI models as scientific Research Assistants (EAIRA) developed at Argonne National Laboratory. This methodology incorporates four primary classes of evaluations. 1) Multiple Choice Questions to assess factual recall; 2) Open Response to evaluate advanced reasoning and problem-solving skills; 3) Lab-Style Experiments involving detailed analysis of capabilities as research assistants in controlled environments; and 4) Field-Style Experiments to capture researcher-LLM interactions at scale in a wide range of scientific domains and applications. These complementary methods enable a comprehensive analysis of LLM strengths and weaknesses with respect to their scientific knowledge, reasoning abilities, and adaptability. Recognizing the rapid pace of LLM advancements, we designed the methodology to evolve and adapt so as to ensure its continued relevance and applicability. This paper describes the methodology state at the end of February 2025. Although developed within a subset of scientific domains, the methodology is designed to be generalizable to a wide range of scientific domains., Comment: 33 pages, 18 figures

This paper develops and implements the stochastic projected Gross-Pitaevskii equation for spin-1 Bose gases, addressing key considerations for numerical simulations. As an application of the theory we explore equilibrium phases in a two-dimensional spin-1 gas, where quasi-long-range order emerges via a Berezinskii-Kosterlitz-Thouless transition. Our analysis includes definition of superfluid densities for both mass and spin degrees of freedom, in a manner suitable for implementation within a stochastic projected Gross-Pitaevskii equation simulation. We present a finite-temperature phase diagram for the ferromagnetic spin-1 Bose gas and identify three distinct superfluid phases: two exhibiting conventional Berezinskii-Kosterlitz-Thouless-like behavior and a novel phase that simultaneously supports independent mass and spin superflows. As temperature increases, the stability region of this novel phase shrinks. This work provides a foundation for further studies of nonequilibrium and finite-temperature phenomena in spinor Bose gases., Comment: 13 pages, 5 figures

We derive sharp upper and lower bounds for the pointwise concentration function of the maximum statistic of $d$ identically distributed real-valued random variables. Our first main result places no restrictions either on the common marginal law of the samples or on the copula describing their joint distribution. We show that, in general, strictly sublinear dependence of the concentration function on the dimension $d$ is not possible. We then introduce a new class of copulas, namely those with a convex diagonal section, and demonstrate that restricting to this class yields a sharper upper bound on the concentration function. This allows us to establish several new dimension-independent and poly-logarithmic-in-$d$ anti-concentration inequalities for a variety of marginal distributions under mild dependence assumptions. Our theory improves upon the best known results in certain special cases. Applications to high-dimensional statistical inference are presented, including a specific example pertaining to Gaussian mixture approximations for factor models, for which our main results lead to superior distributional guarantees., Comment: 22 pages, 2 figures

Single-shot terahertz (THz) spectroscopy probes sub-picosecond, non-repetitive events by combining the advantages of laser absorption techniques with the phase-sensitive detection of interferometric methods. However, its usage as a quantitative tool is hindered by the chirp penalty of the spectral encoding scheme, where a narrow probe bandwidth distorts the THz signal, limiting the bandwidth and spectral resolution of THz measurements. In this work, we introduce Supercontinuum-Enhanced Terahertz Spectroscopy (SETS), a method that leverages a broadband supercontinuum probe to overcome these challenges. With SETS, we show an increase in usable THz bandwidth from 1.5 THz to 2.3 THz, reducing signal distortion by 50% and offering a scalable pathway to extend bandwidth further. Numerical models of spectral encoding highlight the flexibility of SETS, achieving high spectral resolution (< 4 GHz with a 250 ps supercontinuum pulse) while maintaining a usable bandwidth > 1 THz for amplitude and phase spectra. Experiments on argon plasma and water vapor, complemented by theoretical validation, generalization, and extrapolation, show the capability of SETS to measure electron density, collision frequency, and absorption spectra simultaneously with tailored measurement accuracy and resolution (as low as 5x10^15 m^-3 for electron density). By addressing a critical limitation in single-shot diagnostics, SETS enables high-resolution, non-intrusive, quantitative measurements for complex reactive flows and dynamic refractive media.

Measures of textual similarity and divergence are increasingly used to study cultural change. But which measures align, in practice, with social evidence about change? We apply three different representations of text (topic models, document embeddings, and word-level perplexity) to three different corpora (literary studies, economics, and fiction). In every case, works by highly-cited authors and younger authors are textually ahead of the curve. We don't find clear evidence that one representation of text is to be preferred over the others. But alignment with social evidence is strongest when texts are represented through the top quartile of passages, suggesting that a text's impact may depend more on its most forward-looking moments than on sustaining a high level of innovation throughout., Comment: Accepted CHR 2024

Deduplication is a major focus for assembling and curating training datasets for large language models (LLM) -- detecting and eliminating additional instances of the same content -- in large collections of technical documents. Unrestrained, duplicates in the training dataset increase training costs and lead to undesirable properties such as memorization in trained models or cheating on evaluation. Contemporary approaches to document-level deduplication are often extremely expensive in both runtime and memory. We propose LSHBloom, an extension to MinhashLSH, which replaces the expensive LSHIndex with lightweight Bloom filters. LSHBloom demonstrates the same deduplication performance as MinhashLSH with only a marginal increase in false positives (as low as 1e-5 in our experiments); demonstrates competitive runtime (270\% faster than MinhashLSH on peS2o); and, crucially, uses just 0.6\% of the disk space required by MinhashLSH to deduplicate peS2o. We demonstrate that this space advantage scales with increased dataset size -- at the extreme scale of several billion documents, LSHBloom promises a 250\% speedup and a 54$\times$ space advantage over traditional MinHashLSH scaling deduplication of text datasets to many billions of documents.

Many scientific applications opt for particles instead of meshes as their basic primitives to model complex systems composed of billions of discrete entities. Such applications span a diverse array of scientific domains, including molecular dynamics, cosmology, computational fluid dynamics, and geology. The scale of the particles in those scientific applications increases substantially thanks to the ever-increasing computational power in high-performance computing (HPC) platforms. However, the actual gains from such increases are often undercut by obstacles in data management systems related to data storage, transfer, and processing. Lossy compression has been widely recognized as a promising solution to enhance scientific data management systems regarding such challenges, although most existing compression solutions are tailored for Cartesian grids and thus have sub-optimal results on discrete particle data. In this paper, we introduce LCP, an innovative lossy compressor designed for particle datasets, offering superior compression quality and higher speed than existing compression solutions. Specifically, our contribution is threefold. (1) We propose LCP-S, an error-bound aware block-wise spatial compressor to efficiently reduce particle data size. This approach is universally applicable to particle data across various domains. (2) We develop LCP, a hybrid compression solution for multi-frame particle data, featuring dynamic method selection and parameter optimization. (3) We evaluate our solution alongside eight state-of-the-art alternatives on eight real-world particle datasets from seven distinct domains. The results demonstrate that our solution achieves up to 104% improvement in compression ratios and up to 593% increase in speed compared to the second-best option, under the same error criteria., Comment: Accepted by SIGMOD'25

Modern scientific simulations generate massive volumes of data, creating significant challenges for I/O and storage systems. Error-bounded lossy compression (EBLC) offers a solution by reducing dataset sizes while preserving data quality within user-specified limits. This study provides the first comprehensive energy characterization of state-of-the-art EBLC algorithms across various scientific datasets, CPU architectures, and operational modes. We analyze the energy consumption patterns of compression and decompression operations, as well as the energy trade-offs in data I/O scenarios. Our findings demonstrate that EBLC can significantly reduce I/O energy consumption, with savings of up to two orders of magnitude compared to uncompressed I/O for large datasets. In multi-node HPC environments, we observe energy reductions of approximately 25% when using EBLC. We also show that EBLC can achieve compression ratios of 10-100x, potentially reducing storage device requirements by nearly two orders of magnitude. Our work demonstrates the relationships between compression ratios, energy efficiency, and data quality, highlighting the importance of considering compressors and error bounds for specific use cases. Based on our results, we estimate that large-scale HPC facilities could save nearly two orders of magnitude the energy on data writing and significantly reduce storage requirements by integrating EBLC into their I/O subsystems. This work provides a framework for system operators and computational scientists to make informed decisions about implementing EBLC for energy-efficient data management in HPC environments.

GPT-4o is an autoregressive omni model that accepts as input any combination of text, audio, image, and video, and generates any combination of text, audio, and image outputs. It's trained end-to-end across text, vision, and audio, meaning all inputs and outputs are processed by the same neural network. GPT-4o can respond to audio inputs in as little as 232 milliseconds, with an average of 320 milliseconds, which is similar to human response time in conversation. It matches GPT-4 Turbo performance on text in English and code, with significant improvement on text in non-English languages, while also being much faster and 50\% cheaper in the API. GPT-4o is especially better at vision and audio understanding compared to existing models. In line with our commitment to building AI safely and consistent with our voluntary commitments to the White House, we are sharing the GPT-4o System Card, which includes our Preparedness Framework evaluations. In this System Card, we provide a detailed look at GPT-4o's capabilities, limitations, and safety evaluations across multiple categories, focusing on speech-to-speech while also evaluating text and image capabilities, and measures we've implemented to ensure the model is safe and aligned. We also include third-party assessments on dangerous capabilities, as well as discussion of potential societal impacts of GPT-4o's text and vision capabilities.

Accurate characterization of quantum yield is crucial to the reconstruction of energy depositions in silicon at the eV scale. This work presents a new method for experimentally calculating quantum yield using vacuum UV-sensitive silicon photomultipliers (SiPMs), which can be used to determine the probabilities that a UV photon absorbed in a silicon crystal will produce one, two, or three electron-hole pairs. Results are presented which fully constrain the distribution at photon energies up to 7.75eV. This method works by exploiting the saturation of photon detection efficiency which occurs when these devices are biased sufficiently high above their avalanche breakdown voltage. The measured quantum yield values are lower than those that have been previously reported by experimental data and modelling -- this is expected to impact the sensitivity of experiments searching for light dark matter through direct detection in semiconductors, and should also be taken into account when characterizing the performance of UV photodetectors with high quantum efficiency. Measurements have been taken using a Hamamatsu VUV4 and an FBK VUV-HD3 device, showing good agreements between devices, and at a range of temperatures from 163-233K. The validity of the method is assessed using supplementary measurements of absolute photon detection efficiency, and an additional novel method of measuring average quantum yield using DC current-voltage measurements of SiPMs is presented and used for corroboration., Comment: 13 pages, 11 figures

Creative thinking was assessed as an innovative domain for the first time in PISA 2022. This report presents the creative thinking results for Australia as a whole, for the Australian states and territories and for the other groups in PISA 2022. It is one of two thematic reports providing additional high-level analysis of Australia's 2022 PISA performance. Differences in results are only reported if tests of statistical significance showed that these were likely to be real differences; that is, differences were unlikely to be caused by chance.

Kolmogorov-Smirnov (KS) tests rely on the convergence to zero of the KS-distance $d(F_n,G)$ in the one sample case, and of $d(F_n,G_m)$ in the two sample case. In each case the assumption (the null hypothesis) is that $F=G$, and so $d(F,G)=0$. In this paper we extend the Dvoretzky-Kiefer-Wolfowitz-Massart inequality to also apply to cases where $F \neq G$, i.e. when it is possible that $d(F,G) > 0$., Comment: 6 pages, 1 figure

The performance of the GMRES iterative solver on GPUs is limited by the GPU main memory bandwidth. Compressed Basis GMRES outperforms GMRES by storing the Krylov basis in low precision, thereby reducing the memory access. An open question is whether compression techniques that are more sophisticated than casting to low precision can enable large runtime savings while preserving the accuracy of the final results. This paper presents the lightweight in-register compressor FRSZ2 that can decompress at the bandwidth speed of a modern NVIDIA H100 GPU. In an experimental evaluation, we demonstrate using FRSZ2 instead of low precision for compression of the Krylov basis can bring larger runtime benefits without impacting final accuracy.

Error-bounded lossy compression has been a critical technique to significantly reduce the sheer amounts of simulation datasets for high-performance computing (HPC) scientific applications while effectively controlling the data distortion based on user-specified error bound. In many real-world use cases, users must perform computational operations on the compressed data (a.k.a. homomorphic compression). However, none of the existing error-bounded lossy compressors support the homomorphism, inevitably resulting in undesired decompression costs. In this paper, we propose a novel homomorphic error-bounded lossy compressor (called HoSZp), which supports not only error-bounding features but efficient computations (including negation, addition, multiplication, mean, variance, etc.) on the compressed data without the complete decompression step, which is the first attempt to the best of our knowledge. We develop several optimization strategies to maximize the overall compression ratio and execution performance. We evaluate HoSZp compared to other state-of-the-art lossy compressors based on multiple real-world scientific application datasets., Comment: 12 pages, 7 figures, 9 tables

We report on a direct search for elastic photon-photon scattering using x-ray and $\gamma$ photons from a laser-plasma based experiment. A gamma photon beam produced by a laser wakefield accelerator provided a broadband gamma spectrum extending to above $E_\gamma = 200$ MeV. These were collided with a dense x-ray field produced by the emission from a laser heated germanium foil at $E_x \approx 1.4$ keV, corresponding to an invariant mass of $\sqrt{s} = 1.22 \pm 0.22$ MeV. In these asymmetric collisions elastic scattering removes one x-ray and one high-energy $\gamma$ photon and outputs two lower energy $\gamma$ photons. No changes in the $\gamma$ photon spectrum were observed as a result of the collisions allowing us to place a 95% upper bound on the cross section of $1.5 \times 10^{15}\,\mu$b. Although far from the QED prediction, this represents the lowest upper limit obtained so far for $\sqrt{s} \lesssim 1$ MeV., Comment: 9 pages, 7 figures

We show that quantum circuit complexity for the unitary time evolution operator of any time-independent Hamiltonian is bounded by linear growth at early times, independent of any choices of the fundamental gates or cost metric. Deviations from linear early-time growth arise from the commutation algebra of the gates and are manifestly negative for any circuit, decreasing the linear growth rate and leading to a bound on the growth rate of complexity of a circuit at early times. We illustrate this general result by applying it to qubit and harmonic oscillator systems, including the coupled and anharmonic oscillator. By discretizing free and interacting scalar field theories on a lattice, we are also able to extract the early-time behavior and dependence on the lattice spacing of complexity of these field theories in the continuum limit, demonstrating how this approach applies to systems that have been previously difficult to study using existing techniques for quantum circuit complexity., Comment: 32 pages, Version published in JHEP

LLMs have seen rapid adoption in all domains. They need to be trained on high-end high-performance computing (HPC) infrastructures and ingest massive amounts of input data. Unsurprisingly, at such a large scale, unexpected events (e.g., failures of components, instability of the software, undesirable learning patterns, etc.), are frequent and typically impact the training in a negative fashion. Thus, LLMs need to be checkpointed frequently so that they can be rolled back to a stable state and subsequently fine-tuned. However, given the large sizes of LLMs, a straightforward checkpointing solution that directly writes the model parameters and optimizer state to persistent storage (e.g., a parallel file system), incurs significant I/O overheads. To address this challenge, in this paper we study how to reduce the I/O overheads for enabling fast and scalable checkpointing for LLMs that can be applied at high frequency (up to the granularity of individual iterations) without significant impact on the training process. Specifically, we introduce a lazy asynchronous multi-level approach that takes advantage of the fact that the tensors making up the model and optimizer state shards remain immutable for extended periods of time, which makes it possible to copy their content in the background with minimal interference during the training process. We evaluate our approach at scales of up to 180 GPUs using different model sizes, parallelism settings, and checkpointing frequencies. The results show up to 48$\times$ faster checkpointing and 2.2$\times$ faster end-to-end training runtime compared with the state-of-art checkpointing approaches., Comment: Published at HPDC '24: The 33rd International Symposium on High-Performance Parallel and Distributed Computing. Source code at https://github.com/DataStates/datastates-llm

With the STAR experiment at the BNL Relativisic Heavy Ion Collider, we characterize $\sqrt{s_\mathrm{NN}}$ = 200 GeV p+Au collisions by event activity (EA) measured within the pseudorapidity range $eta$ $in$ [-5, -3.4] in the Au-going direction and report correlations between this EA and hard- and soft- scale particle production at midrapidity ($\eta$ $\in$ [-1, 1]). At the soft scale, charged particle production in low-EA p+Au collisions is comparable to that in p+p collisions and increases monotonically with increasing EA. At the hard scale, we report measurements of high transverse momentum (pT) jets in events of different EAs. In contrast with the soft particle production, high-pT particle production and EA are found to be inversely related. To investigate whether this is a signal of jet quenching in high-EA events, we also report ratios of pT imbalance and azimuthal separation of dijets in high- and low-EA events. Within our measurement precision, no significant differences are observed, disfavoring the presence of jet quenching in the highest 30% EA p+Au collisions at $\sqrt{s_\mathrm{NN}}$ = 200 GeV., Comment: 12 page, 9 figures

An easy-plane ferromagnetic spin-1 Bose gas undergoes two Berezinskii-Kosterlitz-Thouless (BKT) transitions, associated with mass and spin superfluidity respectively. We study the effect of axial magnetization on the superfluid properties of this system. We find that nonzero axial magnetization couples mass and spin superflow, via a mechanism analogous to the Andreev-Bashkin effect present in two-component superfluids. With sufficiently large axial magnetization mass and spin superfluidity arise simultaneously. The cross-over to this phase provides a finite-temperature generalization of the zero-temperature broken-axisymmetric to easy-axis transition. We present analytic relations connecting mass and spin superfluidity with experimentally observable coherence of the three spinor components and local magnetization., Comment: 9 pages, 6 figures

Error-bounded lossy compression has been effective in significantly reducing the data storage/transfer burden while preserving the reconstructed data fidelity very well. Many error-bounded lossy compressors have been developed for a wide range of parallel and distributed use cases for years. They are designed with distinct compression models and principles, such that each of them features particular pros and cons. In this paper we provide a comprehensive survey of emerging error-bounded lossy compression techniques. The key contribution is fourfold. (1) We summarize a novel taxonomy of lossy compression into 6 classic models. (2) We provide a comprehensive survey of 10 commonly used compression components/modules. (3) We summarized pros and cons of 46 state-of-the-art lossy compressors and present how state-of-the-art compressors are designed based on different compression techniques. (4) We discuss how customized compressors are designed for specific scientific applications and use-cases. We believe this survey is useful to multiple communities including scientific applications, high-performance computing, lossy compression, and big data., Comment: This paper has been submitted to ACM Computing journal. This is a revised version based on review comments

Learning and Artificial Intelligence (ML/AI) techniques have become increasingly prevalent in high performance computing (HPC). However, these methods depend on vast volumes of floating point data for training and validation which need methods to share the data on a wide area network (WAN) or to transfer it from edge devices to data centers. Data compression can be a solution to these problems, but an in-depth understanding of how lossy compression affects model quality is needed. Prior work largely considers a single application or compression method. We designed a systematic methodology for evaluating data reduction techniques for ML/AI, and we use it to perform a very comprehensive evaluation with 17 data reduction methods on 7 ML/AI applications to show modern lossy compression methods can achieve a 50-100x compression ratio improvement for a 1% or less loss in quality. We identify critical insights that guide the future use and design of lossy compressors for ML/AI., Comment: 12 pages, 4 figures

Large-area next-generation physics experiments rely on using Silicon Photo-Multiplier (SiPM) devices to detect single photons, which trigger charge avalanches. The noise mechanism of external cross-talk occurs when secondary photons produced during a charge avalanche escape from an SiPM and trigger other devices within a detector system. This work presents measured spectra of the secondary photons emitted from the Hamamatsu VUV4 and Fondazione Bruno Kessler VUV-HD3 SiPMs stimulated by laser light, near operational voltages. The work describes the Microscope for the Injection and Emission of Light (MIEL) setup, which is an experimental apparatus constructed for this purpose. Measurements have been performed at a range of over-voltage values and temperatures from 86~K to 293~K. The number of photons produced per avalanche at the source are calculated from the measured spectra and determined to be 40$\pm$9 and 61$\pm$11 photons produced per avalanche for the VUV4 and VUV-HD3 respectively at 4 volts over-voltage. No significant temperature dependence is observed within the measurement uncertainties. The overall number of photons emitted per avalanche from each SiPM device are also reported., Comment: 16 pages, 14 figures

Large Language Models (LLMs) are advancing quickly and impacting people's lives for better or worse. In higher education, concerns have emerged such as students' misuse of LLMs and degraded education outcomes. To unpack the ethical concerns of LLMs for higher education, we conducted a case study consisting of stakeholder interviews (n=20) in higher education computer science. We found that students use several distinct mental models to interact with LLMs - LLMs serve as a tool for (a) writing, (b) coding, and (c) information retrieval, which differ somewhat in ethical considerations. Students and teachers brought up ethical issues that directly impact them, such as inaccurate LLM responses, hallucinations, biases, privacy leakage, and academic integrity issues. Participants emphasized the necessity of guidance and rules for the use of LLMs in higher education, including teaching digital literacy, rethinking education, and having cautious and contextual policies. We reflect on the ethical challenges and propose solutions.