Search

Grokking typically achieves similar loss to ordinary, "steady", learning. We ask whether these different learning paths - grokking versus ordinary training - lead to fundamental differences in the learned models. To do so we compare the features, compressibility, and learning dynamics of models trained via each path in two tasks. We find that grokked and steadily trained models learn the same features, but there can be large differences in the efficiency with which these features are encoded. In particular, we find a novel "compressive regime" of steady training in which there emerges a linear trade-off between model loss and compressibility, and which is absent in grokking. In this regime, we can achieve compression factors 25x times the base model, and 5x times the compression achieved in grokking. We then track how model features and compressibility develop through training. We show that model development in grokking is task-dependent, and that peak compressibility is achieved immediately after the grokking plateau. Finally, novel information-geometric measures are introduced which demonstrate that models undergoing grokking follow a straight path in information space., Comment: Code available at: https://github.com/xand-stapleton/grokking_vs_learning

The interaction of gravitational waves (GWs) with matter is thought to be typically negligible in the Universe. We identify a possible exception in the case of resonant interactions, where GWs emitted by a background binary system such as an inspiraling supermassive black hole (SMBH) binary causes a resonant response in a stellar-mass foreground binary and the frequencies of the two systems become, and remain, synchronized. We point out that such locking is not only possible, but can significantly reduce the binaries' merger time for $\mathcal{O}(1-10^4)$ binaries in the host galaxy of the merging SMBHs of $10^{9-11}M_{\odot}$ for standard general relativity and even more either in ``wet'' SMBH mergers or in certain modified theories of gravity where the inspiral rate is reduced. This could leave an imprint on the period distribution of stellar mass binaries in post-merger galaxies which could be detectable by future GW detectors, such as LISA., Comment: 8 pages, 4 figures, submitted to PRL

Orthogonal time frequency space (OTFS) is a strong candidate waveform for sixth generation wireless communication networks (6G), which can effectively handle time varying wireless channels. In this paper, we analyze the effect of fractional delay in delay Doppler (DD) domain multiplexing techniques. We develop a vector-matrix input-output relationship for the DD domain data transmission system by incorporating the effective pulse shaping filter between the transmitter and receiver along with the channel. Using this input-output relationship, we analyze the effect of the pulse shaping filter on the channel estimation and BER performance in the presence of fractional delay and uncompensated fractional timing offset (TO). For the first time, we propose the use of time-frequency localized (TFL) pulse shaping for the OTFS waveform to overcome the interference due to fractional delays. We show that our proposed TFL-OTFS outperforms the widely used raised cosine pulse-shaped OTFS (RC-OTFS) in the presence of fractional delays. Additionally, TFL-OTFS also shows very high robustness against uncompensated fractional TO, compared to RC-OTFS.

We analyse the stellar distributions on the [Fe/H]-[Mg/Fe] plane for 11 Milky Way-mass galaxies from the FIRE-2 cosmological baryonic zoom-in simulations. Alpha-element bimodality, in the form of two separate sequences on the [Fe/H]-[Mg/Fe] plane, is not a universal feature of disk galaxies. Five galaxies demonstrate double sequences with the $\alpha$-enriched one being older and kinematically hotter, in qualitative agreement with the high-$\alpha$ and low-$\alpha$ populations in the Milky Way disk; three galaxies have unimodal distribution, two show weakly-bimodal features where low-$\alpha$ sequence is visible only over a short range of metallicities, and one show strong bimodality with a different slope of high-$\alpha$ population. We examine the galaxies' gas accretion history over the last 8 Gyr, when bimodal sequences emerge, and demonstrate that the presence of the low-$\alpha$ sequence in the bimodal galaxies is related to the recent infall of metal-poor gas from the circumgalactic medium that joins the galaxy in the outskirts and induces significant growth of the gas disks compared to their non-bimodal counterparts. We also analyse the sources of the accreted gas and illustrate that both gas-rich mergers and smooth accretion of ambient gas can be the source of the accreted gas, and create slightly different bimodal patterns.

In some recent papers, researchers have found two very good methods for reordering columns within supernodes in sparse Cholesky factors; these reorderings can be very useful for certain factorization methods. The first of these reordering methods is based on modeling the underlying problem as a traveling salesman problem (TSP), and the second of these methods is based on partition refinement (PR). In this paper, we devise a fair way to compare the two methods. While the two methods are virtually the same in the quality of the reorderings that they produce, PR should be the method of choice because PR reorderings can be computed using far less time and storage than TSP reorderings.

Detection of the weak cosmological signal from high-redshift hydrogen demands careful data analysis and an understanding of the full instrument signal chain. Here we use the WODEN simulation pipeline to produce realistic data from the Murchison Widefield Array Epoch of Reionisation experiment, and test the effects of different instrumental systematics through the AusEoRPipe analysis pipeline. The simulations include a realistic full sky model, direction-independent calibration, and both random and systematic instrumental effects. Results are compared to matched real observations. We find that, (i) with a sky-based calibration and power spectrum approach we have need to subtract more than 90% of all unresolved point source flux (10mJy apparent) to recover 21-cm signal in the absence of instrumental effects; (ii) when including diffuse emission in simulations, some k-modes cannot be accessed, leading to a need for some diffuse emission removal; (iii) the single greatest cause of leakage is an incomplete sky model; (iv) other sources of errors, such as cable reflections, flagged channels and gain errors, impart comparable systematic power to one another, and less power than the incomplete skymodel., Comment: 13 pages, 11 figures, 1 table Accepted for publication to PASA

With recent advances in Large Language Models (LLMs), Agentic AI has become phenomenal in real-world applications, moving toward multiple LLM-based agents to perceive, learn, reason, and act collaboratively. These LLM-based Multi-Agent Systems (MASs) enable groups of intelligent agents to coordinate and solve complex tasks collectively at scale, transitioning from isolated models to collaboration-centric approaches. This work provides an extensive survey of the collaborative aspect of MASs and introduces an extensible framework to guide future research. Our framework characterizes collaboration mechanisms based on key dimensions: actors (agents involved), types (e.g., cooperation, competition, or coopetition), structures (e.g., peer-to-peer, centralized, or distributed), strategies (e.g., role-based or model-based), and coordination protocols. Through a review of existing methodologies, our findings serve as a foundation for demystifying and advancing LLM-based MASs toward more intelligent and collaborative solutions for complex, real-world use cases. In addition, various applications of MASs across diverse domains, including 5G/6G networks, Industry 5.0, question answering, and social and cultural settings, are also investigated, demonstrating their wider adoption and broader impacts. Finally, we identify key lessons learned, open challenges, and potential research directions of MASs towards artificial collective intelligence.

Single-enzyme catalysis offers a promising approach for unravelling the dynamic behaviour of individual enzymes as they undergo a reaction, revealing the complex heterogeneity that is lost in the averaged ensembles. Here we demonstrate real-time, label-free monitoring of the electrical transduction of single-protein enzymatic activity for two redox enzymes, cytochrome P450cam and glutathione reductase, trapped in an electrochemically controlled nanoscale tunnelling junction immersed in the aqueous enzymatic mixture. The conductance switching signal observed in individual transients of the electrical current flowing through the single-protein junction shows that the tunnelling conductance is modulated by the enzymatic reaction; subtle changes of the enzyme redox state occurring during the chemical catalysis process result in fluctuations of the enzyme junction conductivity, which are captured as a switching signal. At the applied electrochemical reducing potential for electrocatalysis, the transient oxidation of the trapped enzyme in every catalytic cycle opens an additional redox-mediated electron tunnelling channel in the single protein junction that results in a temporary current jump, contributing to the observed conductance switching features. The latter is experimentally assessed via electrochemically controlled conductance measurements of the single-protein junction. The statistical analysis of the switching events captured over long time periods results in average frequencies that correlate well with the reported catalytic turnover values of both enzymes obtained in standard bulk assays. The single-enzyme experiments reveal the acute heterogenous behaviour of enzymatic catalysis and the quantification of single enzyme turnover frequencies.

Negative binomial (NB) regression is a popular method for identifying differentially expressed genes in genomics data, such as bulk and single-cell RNA sequencing data. However, NB regression makes stringent parametric and asymptotic assumptions, which can fail to hold in practice, leading to excess false positive and false negative results. We propose the permuted score test, a new strategy for robust regression based on permuting score test statistics. The permuted score test provably controls type-I error across a much broader range of settings than standard NB regression while nevertheless approximately matching standard NB regression with respect to power (when the assumptions of standard NB regression obtain) and computational efficiency. We accelerate the permuted score test by leveraging emerging techniques for sequential Monte-Carlo testing and novel algorithms for efficiently computing GLM score tests. We apply the permuted score test to real and simulated RNA sequencing data, finding that it substantially improves upon the error control of existing NB regression implementations, including DESeq2. The permuted score test could enhance the reliability of differential expression analysis across diverse biological contexts.

Following the remarkable success of Large Language Models (LLMs) in NLP tasks, there is increasing interest in extending their capabilities to speech -- the most common form of communication. The most widespread approach to integrating speech into LLMs is dense feature prepending (DFP), which prepends the projected speech representations to the textual representations, allowing end-to-end training with a speech encoder. This raises questions about the need for a sophisticated speech encoder for DFP and how its performance compares with a standard encoder-decoder (i.e., cross-attention) architecture. We compare DFP and cross-attention under a variety of configurations, such as CTC compression, sequence-level knowledge distillation, on monolingual, bilingual, and multilingual models. To perform a controlled architectural comparison, we train all models from scratch rather than using large pretrained models and use comparable data and parameter settings, testing speech-to-text recognition (ASR) and translation (ST) on MuST-C v1.0 and CoVoST2 datasets. Despite the wide adoption of DFP, our results do not indicate a clear advantage of DFP over cross-attention., Comment: Accepted at NAACL 2025

The testing and quality assurance of cryogenic superconducting detectors is a time- and labor-intensive process. As experiments deploy increasingly larger arrays of detectors, new methods are needed for performing this testing quickly. Here, we propose a process for flagging under-performing detector wafers before they are ever tested cryogenically. Detectors are imaged under an optical microscope, and computer vision techniques are used to analyze the images, searching for visual defects and other predictors of poor performance. Pipeline performance is verified via a suite of images with simulated defects, yielding a detection accuracy of 98.6%. Lastly, results from running the pipeline on prototype microwave kinetic inductance detectors from the planned SPT-3G+ experiment are presented., Comment: 24 pages, 14 figures. To be submitted to JINST

We examined the anomalies in the light curves of the lensing events MOA-2022-BLG-033, KMT-2023-BLG-0119, and KMT-2023-BLG-1896. We conducted detailed modeling of the light curves to uncover the nature of the anomalies. This modeling revealed that all signals originated from planetary companions to the primary lens. The planet-to-host mass ratios are very low: $q\sim 7.5\times 10^{-5}$ for MOA-2022-BLG-033, $q\sim 3.6\times 10^{-4}$ for KMT-2023-BLG-0119, and $q\sim 6.9\times 10^{-5}$ for KMT-2023-BLG-1896. The anomalies occurred as the source passed through the negative deviation region behind the central caustic along the planet-host axis. The solutions are subject to a common inner-outer degeneracy, resulting in variations in estimating the projected planet-host separation. For KMT-2023-BLG-1896, although the planetary scenario provides the best explanation of the anomaly, the binary companion scenario is marginally possible. We estimate the physical parameters of the planetary systems through Bayesian analyses based on the lensing observables. The analysis identifies MOA-2022-BLG-033L as a planetary system with an ice giant, approximately 12 times the mass of Earth, orbiting an early M dwarf star. The companion of KMT-2023-BLG-1896L is also an ice giant, with a mass around 16 Earth masses, orbiting a mid-K-type main-sequence star. The companion of KMT-2023-BLG-0119L, which has a mass about the mass of Saturn, orbits a mid-K-type dwarf star. The lens for MOA-2022-BLG-033 is highly likely to be located in the disk, whereas for the other events, the probabilities of the lens being in the disk or the bulge are roughly comparable., Comment: 9 pages, 10 figures, 6 tables

The matter power spectrum, $P(k)$, is one of the fundamental quantities in the study of large-scale structure in cosmology. Here, we study its small-scale asymptotic limit, and show that for cold dark matter in $d$ spatial dimensions, $P(k)$ has a universal $k^{-d}$ asymptotic scaling with the wave-number $k$, for $k \gg k_{\rm nl}$, where $k_{\rm nl}^{-1}$ denotes the length scale at which non-linearities in gravitational interactions become important. We propose a theoretical explanation for this scaling, based on a non-perturbative analysis of the system's phase-space structure. Gravitational collapse is shown to drive a turbulent phase-space flow of the quadratic Casimir invariant, where the linear and non-linear time scales are balanced, and this balance dictates the $k$ dependence of the power spectrum. A parallel is drawn to Batchelor turbulence in hydrodynamics, where large scales mix smaller ones via tidal interactions. The $k^{-d}$ scaling is also derived by expressing $P(k)$ as a phase-space integral in the framework of kinetic field theory, which is analysed by the saddle-point method; the dominant critical points of this integral are precisely those where the time scales are balanced. The coldness of the dark-matter distribution function -- its non-vanishing only on a $d$-dimensional sub-manifold of phase-space -- underpins both approaches. The theory is accompanied by $1\mathrm{D}$ Vlasov--Poisson simulations, which confirm it., Comment: Submitted, comments welcome

Information theory has provided foundations for the theories of several application areas critical for modern society, including communications, computer storage, and AI. A key aspect of Shannon's 1948 theory is a sharp lower bound on the number of bits needed to encode and communicate a string of symbols. When he introduced the theory, Shannon famously excluded any notion of semantics behind the symbols being communicated. This semantics-free notion went on to have massive impact on communication and computing technologies, even as multiple proposals for reintroducing semantics in a theory of information were being made, notably one where Carnap and Bar-Hillel used logic and reasoning to capture semantics. In this paper we present, for the first time, a Shannon-style analysis of a communication system equipped with a deductive reasoning capability, implemented using logical inference. We use some of the most important techniques developed in information theory to demonstrate significant and sometimes surprising gains in communication efficiency availed to us through such capability, demonstrated also through practical codes. We thus argue that proposals for a semantic information theory should include the power of deductive reasoning to magnify the value of transmitted bits as we strive to fully unlock the inherent potential of semantics.

Laser-based displays are highly sought after for their superior brightness and color performance, especially in advanced applications like augmented reality (AR). However, their broader adoption has been hindered by bulky projector designs and complex optical module assemblies. Here, we introduce a new laser display architecture enabled by large-scale visible photonic integrated circuits (PICs) to address these challenges. Unlike previous projector-style laser displays, this architecture features an ultra-thin, flat-panel form factor, replacing bulky free-space illumination modules with a single, high-performance photonic chip. Centimeter-scale PIC devices, which integrate thousands of distinct optical components on-chip, are carefully tailored to achieve high display uniformity, contrast, and efficiency. We demonstrate a 2 mm-thick flat-panel laser display combining the PIC with a liquid-crystal-on-silicon (LCoS) panel, achieving 211% of the color gamut and more than 80% volume reduction compared to traditional LCoS displays. We further showcase its application in a see-through AR system. Our work represents a major advancement in the integration of nanophotonics with display technology, enabling a range of new display concepts, from high-performance immersive displays to slim-panel 3D holography.

Large Multimodal Models (LMMs) have demonstrated impressive performance across numerous academic benchmarks. However, fine-tuning still remains essential to achieve satisfactory performance on downstream tasks, while the task-specific tuning samples are usually not readily available or expensive and time-consuming to obtain. To address this, we propose an error-driven data-efficient tuning framework that aims to efficiently adapt generic LMMs to newly emerging tasks without requiring any task-specific training samples. In our approach, a generic LMM, acting as a student model, is first evaluated on a small validation set of the target task, and then a more powerful model, acting as a teacher model, identifies the erroneous steps within the student model's reasoning steps and analyzes its capability gaps from fully addressing the target task. Based on these gaps, targeted training samples are further retrieved from existing task-agnostic datasets to tune the student model and tailor it to the target task. We perform extensive experiments across three different training data scales and seven tasks, demonstrating that our training paradigm significantly and efficiently improves LMM's performance on downstream tasks, achieving an average performance boost of 7.01%., Comment: 16 pages, 6 figures

This summary of the second Terrestrial Very-Long-Baseline Atom Interferometry (TVLBAI) Workshop provides a comprehensive overview of our meeting held in London in April 2024, building on the initial discussions during the inaugural workshop held at CERN in March 2023. Like the summary of the first workshop, this document records a critical milestone for the international atom interferometry community. It documents our concerted efforts to evaluate progress, address emerging challenges, and refine strategic directions for future large-scale atom interferometry projects. Our commitment to collaboration is manifested by the integration of diverse expertise and the coordination of international resources, all aimed at advancing the frontiers of atom interferometry physics and technology, as set out in a Memorandum of Understanding signed by over 50 institutions., Comment: Summary of the second Terrestrial Very-Long-Baseline Atom Interferometry Workshop held at Imperial College London: https://indico.cern.ch/event/1369392/

The density-density response in optimally doped Bi$_2$Sr$_2$CaCu$_2$O$_{8+x}$ has recently been shown to exhibit conformal symmetry. Using, the experimentally inferred conformal dynamic susceptibility, we compute the resultant quantum Fisher information (QFI), a witness to multi-partite entanglement. In contrast to a Fermi liquid in which the QFI is approximately temperature independent much below the Fermi energy scale, we find that the QFI increases as a power law at low temperatures but ultimately extrapolates to a constant at $T=0$. The constant is of the form, $\omega_g^{2\Delta}$, where $\Delta$ is the conformal dimension and $\omega_g$ is the UV cutoff which is on the order of the pseudogap. As this constant {depends on both UV and IR properties}, it illustrates that multipartite entanglement in a strange metal exhibits UV-IR mixing, a benchmark feature of doped Mott insulators as exemplified by dynamical spectral weight transfer. We conclude with a discussion of the implication of our results for low-energy reductions of the Hubbard model., Comment: For an issue of Physica C dedicated to Jan Zaanen

Generic sentences express generalisations about the world without explicit quantification. Although generics are central to everyday communication, building a precise semantic framework has proven difficult, in part because speakers use generics to generalise properties with widely different statistical prevalence. In this work, we study the implicit quantification and context-sensitivity of generics by leveraging language models as models of language. We create ConGen, a dataset of 2873 naturally occurring generic and quantified sentences in context, and define p-acceptability, a metric based on surprisal that is sensitive to quantification. Our experiments show generics are more context-sensitive than determiner quantifiers and about 20% of naturally occurring generics we analyze express weak generalisations. We also explore how human biases in stereotypes can be observed in language models., Comment: Accepted at CoLing 2025

Gravitational-wave observatories like LIGO are large-scale, terrestrial instruments housed in infrastructure that spans a multi-kilometer geographic area and which must be actively controlled to maintain operational stability for long observation periods. Despite exquisite seismic isolation, they remain susceptible to seismic noise and other terrestrial disturbances that can couple undesirable vibrations into the instrumental infrastructure, potentially leading to control instabilities or noise artifacts in the detector output. It is, therefore, critical to characterize the seismic state of these observatories to identify a set of temporal patterns that can inform the detector operators in day-to-day monitoring and diagnostics. On a day-to-day basis, the operators monitor several seismically relevant data streams to diagnose operational instabilities and sources of noise using some simple empirically-determined thresholds. It can be untenable for a human operator to monitor multiple data streams in this manual fashion and thus a distillation of these data-streams into a more human-friendly format is sought. In this paper, we present an end-to-end machine learning pipeline for features-based multivariate time series clustering to achieve this goal and to provide actionable insights to the detector operators by correlating found clusters with events of interest in the detector., Comment: 8 pages, 6 figures, Accepted to The 5th International Workshop on Big Data & AI Tools, Methods, and Use Cases for Innovative Scientific Discovery (BTSD 2024)

Given recent increases in ocean water levels brought on by climate change, this investigation decomposed changes in coastal water levels into its fundamental components to predict maximum water levels for a given coastal location. The study focused on Virginia Key, Florida, in the United States, located near the coast of Miami. Hourly mean lower low water (MLLW) levels were obtained from the National Data Buoy Center from January 28, 1994, through December 31, 2023. In the temporal dimension, Kolmogorov-Zurbenko filters were used to extract long-term trends, annual and daily tides, and higher frequency harmonics, while in the spectral dimension, Kolmogorov-Zurbenko periodograms with DiRienzo-Zurbenko algorithm smoothing were used to confirm known tidal frequencies and periods. A linear model predicted that the long-term trend in water level will rise 2.02 feet from January 1994 to December 2050, while a quadratic model predicted a rise of 5.91 during the same period. In addition, the combined crests of annual tides, daily tides, and higher frequency harmonics increase water levels up to 2.16 feet, yielding a combined total of 4.18 feet as a lower bound and a combined total of 8.09 feet as an upper bound. These findings provide a foundation for more accurate prediction of coastal flooding during severe weather events and provide an impetus for policy choices with respect to residential communities, businesses, and wildlife habitats. Further, using Kolmogorov-Zurbenko analytic methods to study coastal sites throughout the world could draw a more comprehensive picture of the impact climate change is having on coastal waters globally., Comment: 23 pages, 7 figures, this article draws from arXiv:2007.03031v3 and arXiv:2412.07735v1

Both encoder-only models (e.g., BERT, RoBERTa) and large language models (LLMs, e.g., Llama3) have been widely used for text classification tasks. However, there is a lack of systematic studies comparing the performance of encoder-based models and LLMs in text classification, particularly when fine-tuning is involved. This study employed a diverse range of models and methods, varying in size and architecture, and including both fine-tuned and pre-trained approaches. We first assessed the performances of these LLMs on the 20 Newsgroups (20NG) and MASSIVE datasets, comparing them to encoder-only RoBERTa models. Additionally, we explored the multi-task capabilities of both model types by combining multiple classification tasks, including intent detection and slot-filling, into a single model using data from both datasets. Our results indicate that fully fine-tuned Llama3-70B models outperform RoBERTa-large and other decoder LLMs across various classification tasks and datasets. Moreover, the consolidated multi-task fine-tuned LLMs matched the performance of dual-model setups in both tasks across both datasets. Overall, our study provides a comprehensive benchmark of encoder-only and LLM models on text classification tasks and demonstrates a method to combine two or more fully fine-tuned decoder LLMs for reduced latency and equivalent performance., Comment: 9 pages, 3 tables

Confining electrons or holes in quantum dots formed in the channel of industry-standard fully depleted silicon-on-insulator CMOS structures is a promising approach to scalable qubit architectures. In this communication, we present measurement results of a commercial nanostructure fabricated using the GlobalFoundries 22FDX(TM) industrial process. We demonstrate here that quantum dots are formed in the device channel by applying a combination of a back- and gate voltages. We report our results on an effective detuning of the energy levels in the quantum dots by varying the barrier gate voltages in combination with the back-gate voltage. Given the need and importance of scaling to larger numbers of qubits, we demonstrate here the feasibility of single-electron box sensors at the edge of the quantum dot array for effective charge sensing in different operation modes -- sensing charge transitions in a single- and double quantum dots forming the quantum dot array. We also report measurement results demonstrating bias triangle pair formation and precise control over coupled quantum dots with variations in the inter-dot barrier. The reported measurement results demonstrate the ability to control the formation and coupling of multiple quantum dots in a quantum dot array and to sense their charge state via a Single Electron Box sensor in a commercial process for the first time.

This investigation establishes the theoretical and practical limits of signal strength estimate precision for Kolmogorov-Zurbenko periodograms with dynamic smoothing and compares them to those of standard log-periodograms with static smoothing. Previous research has established the sensitivity, accuracy, resolution, and robustness of Kolmogorov-Zurbenko periodograms with dynamic smoothing in estimating signal frequencies. However, the precision with which they estimate signal strength has never been evaluated. To this point, the width of the confidence interval for a signal strength estimate can serve as a criterion for assessing the precision of such estimates: the narrower the confidence interval, the more precise the estimate. The statistical background for confidence intervals of periodograms is presented, followed by candidate functions to compute and plot them when using Kolmogorov-Zurbenko periodograms with dynamic smoothing. Given an identified signal frequency, a static smoothing window and its smoothing window width can be selected such that its confidence interval is narrower and, thus, its signal strength estimate more precise, than that of dynamic smoothing windows, all while maintaining a level of frequency resolution as good as or better than that of a dynamic smoothing window. These findings suggest the need for a two-step protocol in spectral analysis: computation of a Kolmogorov-Zurbenko periodogram with dynamic smoothing to detect, identify, and separate signal frequencies, followed by computation of a Kolmogorov-Zurbenko periodogram with static smoothing to precisely estimate signal strength and compute its confidence intervals., Comment: 32 pages, 8 figures, this article draws from arXiv:2007.03031v3

The growing demand for accurate, continuous, and non-invasive health monitoring has propelled multi-sensor data fusion to the forefront of healthcare technology. This review aims to provide an overview of the development of fusion frameworks in the literature and common terminology used in fusion literature. The review introduces the fusion classification standards and methods that are most relevant from an algorithm development perspective. Applications of the reviewed fusion frameworks in fields such as defense, autonomous driving, robotics, and image fusion are also discussed to provide contextual information on the various fusion methodologies that have been developed in this field. This review provides a comprehensive analysis of multi-sensor data fusion methods applied to health monitoring systems, focusing on key algorithms, applications, challenges, and future directions. We examine commonly used fusion techniques, including Kalman filters, Bayesian networks, and machine learning models. By integrating data from various sources, these fusion approaches enhance the reliability, accuracy, and resilience of health monitoring systems. However, challenges such as data quality and differences in acquisition systems exist, calling for intelligent fusion algorithms in recent years. The review finally converges on applications of fusion algorithms in biomedical inference tasks like heartbeat detection, respiration rate estimation, sleep apnea detection, arrhythmia detection, and atrial fibrillation detection.

We define the triple Riordan group, whose elements consist of $4$-tuples of power series $(g, f_1, f_2, f_3)$ with $g\in \mathbf{R}[[x^3]]$, and $f_1, f_2, f_3 \in x\mathbf{R}[[x^3]]$, for an appropriate ring $\mathbf{R}$. The construction of this group generalizes that of the double Riordan group, and lays the pattern for further generalizations., Comment: 10 pages

Table extraction from document images is a challenging AI problem, and labelled data for many content domains is difficult to come by. Existing table extraction datasets often focus on scientific tables due to the vast amount of academic articles that are readily available, along with their source code. However, there are significant layout and typographical differences between tables found across scientific, financial, and other domains. Current datasets often lack the words, and their positions, contained within the tables, instead relying on unreliable OCR to extract these features for training modern machine learning models on natural language processing tasks. Therefore, there is a need for a more general method of obtaining labelled data. We present SynFinTabs, a large-scale, labelled dataset of synthetic financial tables. Our hope is that our method of generating these synthetic tables is transferable to other domains. To demonstrate the effectiveness of our dataset in training models to extract information from table images, we create FinTabQA, a layout large language model trained on an extractive question-answering task. We test our model using real-world financial tables and compare it to a state-of-the-art generative model and discuss the results. We make the dataset, model, and dataset generation code publicly available., Comment: 12 pages, 8 figures

Large language models (LLMs) are increasingly being adapted to achieve task-specificity for deployment in real-world decision systems. Several previous works have investigated the bias transfer hypothesis (BTH) by studying the effect of the fine-tuning adaptation strategy on model fairness to find that fairness in pre-trained masked language models have limited effect on the fairness of models when adapted using fine-tuning. In this work, we expand the study of BTH to causal models under prompt adaptations, as prompting is an accessible, and compute-efficient way to deploy models in real-world systems. In contrast to previous works, we establish that intrinsic biases in pre-trained Mistral, Falcon and Llama models are strongly correlated (rho >= 0.94) with biases when the same models are zero- and few-shot prompted, using a pronoun co-reference resolution task. Further, we find that bias transfer remains strongly correlated even when LLMs are specifically prompted to exhibit fair or biased behavior (rho >= 0.92), and few-shot length and stereotypical composition are varied (rho >= 0.97). Our findings highlight the importance of ensuring fairness in pre-trained LLMs, especially when they are later used to perform downstream tasks via prompt adaptation.

This paper presents a characterization of a two-qubit processor in a 6-quantum dot array in SiGe, from the perspective of its quantum information processing capabilities. The analysis includes randomized benchmarking of single- and two-qubit gates, SPAM characterization, and Bell's state tomography; all basic functionality required for universal quantum computation. In light of our efforts to combine spin qubits with integrated cryogenic electronics, we evaluate the qubits' performance metrics at 300mK and 740mK. The latter temperature lies within the realistic thermal budget for integrated cryogenic electronics, making it particularly relevant for assessing qubit performance in practical scenarios., Comment: 13 pages, 7 figures, 4 tables

We present the Materials Learning Algorithms (MALA) package, a scalable machine learning framework designed to accelerate density functional theory (DFT) calculations suitable for large-scale atomistic simulations. Using local descriptors of the atomic environment, MALA models efficiently predict key electronic observables, including local density of states, electronic density, density of states, and total energy. The package integrates data sampling, model training and scalable inference into a unified library, while ensuring compatibility with standard DFT and molecular dynamics codes. We demonstrate MALA's capabilities with examples including boron clusters, aluminum across its solid-liquid phase boundary, and predicting the electronic structure of a stacking fault in a large beryllium slab. Scaling analyses reveal MALA's computational efficiency and identify bottlenecks for future optimization. With its ability to model electronic structures at scales far beyond standard DFT, MALA is well suited for modeling complex material systems, making it a versatile tool for advanced materials research.