Search

Maxwell's demon (MD) has proven an instructive vehicle by which to explore the relationship between information theory and thermodynamics, fueling the possibility of information driven machines. A long standing debate has been the concern of entropy violation, now resolved by the introduction of a quantum MD, but this theoretical suggestion has proven experimentally challenging. Here, we use classical vectorially structured light that is non-separable in spin and orbital angular momentum to emulate a quantum MD experiment. Our classically entangled light fields have all the salient properties necessary of their quantum counterparts but without the experimental complexity of controlling quantum entangled states. We use our experiment to show that the demon's entropy increases during the process while the system's entropy decreases, so that the total entropy is conserved through an exchange of information, confirming the theoretical prediction. We show that our MD is able to extract useful work from the system in the form of orbital angular momentum, opening a path to information driven optical spanners for the mechanical rotation of objects with light. Our synthetic dimensions of angular momentum can easily be extrapolated to other degrees of freedom, for scalable and robust implementations of MDs at both the classical and quantum realms, enlightening the role of a structured light MD and its capability to control and measure information., Comment: 14 pages, 6 Figures

The eXTP mission is a major project of the Chinese Academy of Sciences (CAS), with a large involvement of Europe. Its scientific payload includes four instruments: SFA, PFA, LAD and WFM. They offer an unprecedented simultaneous wide-band Xray timing and polarimetry sensitivity. A large European consortium is contributing to the eXTP study, both for the science and the instrumentation. Europe is expected to provide two of the four instruments: LAD and WFM; the LAD is led by Italy and the WFM by Spain. The WFM for eXTP is based on the design originally proposed for the LOFT ESA M3 mission, that underwent a Phase A feasibility study. It will be a wide field of view X-ray monitor instrument working in the 2-50 keV energy range, achieved with large-area Silicon Drift Detectors (SDDs), similar to the ones used for the LAD but with better spatial resolution. The WFM will consist of 3 pairs of coded mask cameras with a total combined field of view (FoV) of 90x180 degrees at zero response and a source localisation accuracy of ~1 arc min. The main goal of the WFM is to provide triggers for the target of opportunity observations of the SFA, PFA and LAD, in order to perform the core science programme, dedicated to the study of matter under extreme conditions of density, gravity and magnetism. In addition, the unprecedented combination of large field of view and imaging capability, down to 2 keV, of the WFM will allow eXTP to make important discoveries of the variable and transient X-ray sky, and provide X-ray coverage of a broad range of astrophysical objects covered under 'observatory science', such as gamma-ray bursts, fast radio bursts, gravitational wave electromagnetic counterparts. In this paper we provide an overview of the WFM instrument, explaining its design, configuration, and anticipated performance., Comment: 15 pages, 13 figures, Proceedings of SPIE 13093, Space Telescopes and Instrumentation 2024: Ultraviolet to Gamma Ray; Proceedings Volume 13093, Space Telescopes and Instrumentation 2024: Ultraviolet to Gamma Ray; 130931Y (2024); doi: 10.1117/12.3020020

The accurate simulation of complex biochemical phenomena has historically been hampered by the computational requirements of high-fidelity molecular-modeling techniques. Quantum mechanical methods, such as ab initio wave-function (WF) theory, deliver the desired accuracy, but have impractical scaling for modelling biosystems with thousands of atoms. Combining molecular fragmentation with MP2 perturbation theory, this study presents an innovative approach that enables biomolecular-scale ab initio molecular dynamics (AIMD) simulations at WF theory level. Leveraging the resolution-of-the-identity approximation for Hartree-Fock and MP2 gradients, our approach eliminates computationally intensive four-center integrals and their gradients, while achieving near-peak performance on modern GPU architectures. The introduction of asynchronous time steps minimizes time step latency, overlapping computational phases and effectively mitigating load imbalances. Utilizing up to 9,400 nodes of Frontier and achieving 59% (1006.7 PFLOP/s) of its double-precision floating-point peak, our method enables us to break the million-electron and 1 EFLOP/s barriers for AIMD simulations with quantum accuracy.

The design and control of atomic-scale spin structures constitute major challenges for spin-based quantum technology platforms, including quantum dots, color centers, and molecular spins. Here, we showcase a strategy for designing the quantum properties of molecular spin qubits by combining tip-assisted on-surface assembly with electron spin resonance scanning tunneling microscopy (ESR-STM): We fabricate magnetic dimer complexes that consist of an iron phthalocyanine (FePc) molecule and an organometallic half-sandwich complex formed by the FePc ligand and an attached iron atom, Fe(C6H6). The total complex forms a mixed-spin (1/2,1) quantum ferrimagnet with a well-separated correlated ground state doublet, which we utilize for coherent control. As a result of the correlation, the quantum ferrimagnet shows an improved spin lifetime (>1.5 ${\mu}$s) as it is partially protected against inelastic electron scattering. Lastly, the ferrimagnet units also enable intermolecular coupling, that can be used to realize both ferromagnetic or antiferromagnetic structures. Thus, quantum ferrimagnets provide a versatile platform to improve coherent control in general and to study complex magnetic interactions., Comment: 17 pages, 4 figures

Listeners of long-form talk-audio content, such as podcast episodes, often find it challenging to understand the overall structure and locate relevant sections. A practical solution is to divide episodes into chapters--semantically coherent segments labeled with titles and timestamps. Since most episodes on our platform at Spotify currently lack creator-provided chapters, automating the creation of chapters is essential. Scaling the chapterization of podcast episodes presents unique challenges. First, episodes tend to be less structured than written texts, featuring spontaneous discussions with nuanced transitions. Second, the transcripts are usually lengthy, averaging about 16,000 tokens, which necessitates efficient processing that can preserve context. To address these challenges, we introduce PODTILE, a fine-tuned encoder-decoder transformer to segment conversational data. The model simultaneously generates chapter transitions and titles for the input transcript. To preserve context, each input text is augmented with global context, including the episode's title, description, and previous chapter titles. In our intrinsic evaluation, PODTILE achieved an 11% improvement in ROUGE score over the strongest baseline. Additionally, we provide insights into the practical benefits of auto-generated chapters for listeners navigating episode content. Our findings indicate that auto-generated chapters serve as a useful tool for engaging with less popular podcasts. Finally, we present empirical evidence that using chapter titles can enhance effectiveness of sparse retrieval in search tasks., Comment: 9 pages, 4 figures, CIKM industry track 2024

We study agents acting in an unknown environment where the agent's goal is to find a robust policy. We consider robust policies as policies that achieve high cumulative rewards for all possible environments. To this end, we consider agents minimizing the maximum regret over different environment parameters, leading to the study of minimax regret. This research focuses on deriving information-theoretic bounds for minimax regret in Markov Decision Processes (MDPs) with a finite time horizon. Building on concepts from supervised learning, such as minimum excess risk (MER) and minimax excess risk, we use recent bounds on the Bayesian regret to derive minimax regret bounds. Specifically, we establish minimax theorems and use bounds on the Bayesian regret to perform minimax regret analysis using these minimax theorems. Our contributions include defining a suitable minimax regret in the context of MDPs, finding information-theoretic bounds for it, and applying these bounds in various scenarios.

Segmentation and classification of large numbers of instances, such as cell nuclei, are crucial tasks in digital pathology for accurate diagnosis. However, the availability of high-quality datasets for deep learning methods is often limited due to the complexity of the annotation process. In this work, we investigate the impact of noisy annotations on the training and performance of a state-of-the-art CNN model for the combined task of detecting, segmenting and classifying nuclei in histopathology images. In this context, we investigate the conditions for determining an appropriate number of training epochs to prevent overfitting to annotation noise during training. Our results indicate that the utilisation of a small, correctly annotated validation set is instrumental in avoiding overfitting and maintaining model performance to a large extent. Additionally, our findings underscore the beneficial role of pre-training.

The Wide Field Monitor (WFM) is one of the three instruments on the Spectroscopic Time-Resolving Observatory for Broadband Energy X-rays (STROBE-X) mission, which was proposed in response to the NASA 2023 call for a probe class mission. The WFM is a coded-mask camera system that would be the most scientifically capable wide-angle monitor ever flown. The field of view covers one third of the sky, to 50 percent mask coding, and the energy sensitivity is 2 to 50 keV. The WFM is designed to identify new X-ray transients and to capture spectral and timing changes in known sources with data of unprecedented quality. Science applications cover diverse classes, in including X-ray bursts that coincide with gravitational wave detections, gamma ray bursts and their transition from prompt emission to afterglow, subluminous GRBs that may signal shock breakout in supernovae, state transitions in accreting compact objects and their jets, bright flares in fast X-ray transients, accretion onset in transitional pulsars, and coronal flares from many types of active stars., Comment: 19 pages, 5 figures, accepted for publication in JATIS

The Duke Robotics Club is proud to present our robot for the 2024 RoboSub Competition: Oogway. Now in its second year, Oogway has been dramatically upgraded in both its capabilities and reliability. Oogway was built on the principle of independent, well-integrated, and reliable subsystems. Individual components and subsystems were tested and designed separately. Oogway's most advanced capabilities are a result of the tight integration between these subsystems. Such examples include a re-envisioned controls system, an entirely new electrical stack, advanced sonar integration, additional cameras and system monitoring, a new marker dropper, and a watertight capsule mechanism. These additions enabled Oogway to prequalify for Robosub 2024.

SDO-FM is a foundation model using data from NASA's Solar Dynamics Observatory (SDO) spacecraft; integrating three separate instruments to encapsulate the Sun's complex physical interactions into a multi-modal embedding space. This model can be used to streamline scientific investigations involving SDO by making the enormous datasets more computationally accessible for heliophysics research and enable investigations that require instrument fusion. We discuss four key components: an ingestion pipeline to create machine learning ready datasets, the model architecture and training approach, resultant embeddings and fine-tunable models, and finally downstream fine-tuned applications. A key component of this effort has been to include subject matter specialists at each stage of development; reviewing the scientific value and providing guidance for model architecture, dataset, and training paradigm decisions. This paper marks release of our pretrained models and embedding datasets, available to the community on Hugging Face and sdofm.org.

A rising interest in the modality extension of foundation language models warrants discussion on the most effective, and efficient, multimodal training approach. This work focuses on neural machine translation (NMT) and proposes a joint multimodal training regime of Speech-LLM to include automatic speech translation (AST). We investigate two different foundation model architectures, decoder-only GPT and encoder-decoder T5, extended with Canary-1B's speech encoder. To handle joint multimodal training, we propose a novel training framework called EMMeTT. EMMeTT improves training efficiency with the following: balanced sampling across languages, datasets, and modalities; efficient sequential data iteration; and a novel 2D bucketing scheme for multimodal data, complemented by a batch size optimizer (OOMptimizer). We show that a multimodal training consistently helps with both architectures. Moreover, SALM-T5 trained with EMMeTT retains the original NMT capability while outperforming AST baselines on four-language subsets of FLORES and FLEURS. The resultant Multimodal Translation Model produces strong text and speech translation results at the same time., Comment: 4 pages, submitted to ICASSP 2025

We establish a structure theorem, analogous to the classical result of Milnor and Moore, for differential graded Hopf algebras: any differential Hopf algebra $H$ that is free as a coalgebra carries an underlying $B_\infty$ algebra structure that restricts to the subspace of primitives, and conversely $H$ may be recovered via a universal enveloping differential-2-associative algebra. This extends the work of Loday and Ronco [12] where the ungraded non-differential case was treated, and only the multibrace part of the $B_\infty$ structure was found. We show that the multibrace structure of [12] originates from a twisting of a quasi-trivial structure, extending the work of Markl [14] on the $A_\infty$ structure underlying any algebra with a square-zero endomorphism. In this framework it is also clear that the multibrace and $A_\infty$ structures are compatible, and provide an appropriate $B_\infty$ structure for the structure theorem., Comment: 18pp

Quantum computers are promising tools for the simulation of many-body systems, and among those, QCD stands out by its rich phenomenology. Every simulation starts with a codification, and here we succently review a newly developed compact encoding based on the identification between registers and particles; the quantum memory is divided into registers, and to each we associate a Hilbert space of dimension the number of degrees of freedom of the codified particles. In this way we gain an exponential compression over direct encodings for a low number of particles with many degrees of freedom. As an example we apply this encoding on a two-register memory and implement antisymmetrization and exponentiation algorithms., Comment: Proceedings of the 27th High-Energy Physics International Conference in Quantum Chromodynamis (QCD24)

We establish a Crapo complementation formula for the M\"obius function $\mu^X$ in a general decomposition space $X$ in terms of a convex subspace $K$ and its complement: $\mu^X \simeq \mu^{X\setminus K} + \mu^X*\zeta^K*\mu^X$. We work at the objective level, meaning that the formula is an explicit homotopy equivalence of $\infty$-groupoids. Almost all arguments are formulated in terms of (homotopy) pullbacks. Under suitable finiteness conditions on $X$, one can take homotopy cardinality to obtain a formula in the incidence algebra at the level of $\mathbb{Q}$-algebras. When $X$ is the nerve of a locally finite poset, this recovers the Bj\"orner--Walker formula, which in turn specialises to the original Crapo complementation formula when the poset is a finite lattice. A substantial part of the work is to introduce and develop the notion of convexity for decomposition spaces, which in turn requires some general preparation in decomposition-space theory, notably some results on reduced covers and ikeo and semi-ikeo maps. These results may be of wider interest. Once this is set up, the objective proof of the Crapo formula is quite similar to that of Bj\"orner--Walker., Comment: 24pp

As part of our comprehensive, ongoing characterisation of the low-mass end of the main sequence in the Solar neighbourhood, we used the OSIRIS instrument at the 10.4 m Gran Telescopio Canarias to acquire low- and mid-resolution (R$\approx$300 and R$\approx$2500) optical spectroscopy of 53 late-M and L ultracool dwarfs. Most of these objects are known but poorly investigated and lacking complete kinematics. We measured spectral indices, determined spectral types (six of which are new) and inferred effective temperature and surface gravity from BT-Settl synthetic spectra fits for all objects. We were able to measure radial velocities via line centre fitting and cross correlation for 46 objects, 29 of which lacked previous radial velocity measurements. Using these radial velocities in combination with the latest Gaia DR3 data, we also calculated Galactocentric space velocities. From their kinematics, we identified two candidates outside of the thin disc and four in young stellar kinematic groups. Two further ultracool dwarfs are apparently young field objects: 2MASSW J1246467$+$402715 (L4$\beta$), which has a potential, weak lithium absorption line, and G 196$-$3B (L3$\beta$), which was already known as young due to its well-studied primary companion., Comment: 33 pages, 14 figures, Accepted by MNRAS

Understanding space weather is vital for the protection of our terrestrial and space infrastructure. In order to predict space weather accurately, large amounts of data are required, particularly in the extreme ultraviolet (EUV) spectrum. An exquisite source of information for such data is provided by the Solar Dynamic Observatory (SDO), which has been gathering solar measurements for the past 13 years. However, after a malfunction in 2014 affecting the onboard Multiple EUV Grating Spectrograph A (MEGS-A) instrument, the scientific output in terms of EUV measurements has been significantly degraded. Building upon existing research, we propose to utilize deep learning for the virtualization of the defective instrument. Our architecture features a linear component and a convolutional neural network (CNN) -- with EfficientNet as a backbone. The architecture utilizes as input grayscale images of the Sun at multiple frequencies -- provided by the Atmospheric Imaging Assembly (AIA) -- as well as solar magnetograms produced by the Helioseismic and Magnetic Imager (HMI). Our findings highlight how AIA data are all that is needed for accurate predictions of solar irradiance. Additionally, our model constitutes an improvement with respect to the state-of-the-art in the field, further promoting the idea of deep learning as a viable option for the virtualization of scientific instruments., Comment: Submission to Machine Learning and the Physical Sciences Workshop, NeurIPS 2023

In this article, we study the fundamental limits in the design of fair and/or private representations achieving perfect demographic parity and/or perfect privacy through the lens of information theory. More precisely, given some useful data $X$ that we wish to employ to solve a task $T$, we consider the design of a representation $Y$ that has no information of some sensitive attribute or secret $S$, that is, such that $I(Y;S) = 0$. We consider two scenarios. First, we consider a design desiderata where we want to maximize the information $I(Y;T)$ that the representation contains about the task, while constraining the level of compression (or encoding rate), that is, ensuring that $I(Y;X) \leq r$. Second, inspired by the Conditional Fairness Bottleneck problem, we consider a design desiderata where we want to maximize the information $I(Y;T|S)$ that the representation contains about the task which is not shared by the sensitive attribute or secret, while constraining the amount of irrelevant information, that is, ensuring that $I(Y;X|T,S) \leq r$. In both cases, we employ extended versions of the Functional Representation Lemma and the Strong Functional Representation Lemma and study the tightness of the obtained bounds. Every result here can also be interpreted as a coding with perfect privacy problem by considering the sensitive attribute as a secret.

We investigate the in-distribution generalization of machine learning algorithms. We depart from traditional complexity-based approaches by analyzing information-theoretic bounds that quantify the dependence between a learning algorithm and the training data. We consider two categories of generalization guarantees: 1) Guarantees in expectation: These bounds measure performance in the average case. Here, the dependence between the algorithm and the data is often captured by information measures. While these measures offer an intuitive interpretation, they overlook the geometry of the algorithm's hypothesis class. Here, we introduce bounds using the Wasserstein distance to incorporate geometry, and a structured, systematic method to derive bounds capturing the dependence between the algorithm and an individual datum, and between the algorithm and subsets of the training data. 2) PAC-Bayesian guarantees: These bounds measure the performance level with high probability. Here, the dependence between the algorithm and the data is often measured by the relative entropy. We establish connections between the Seeger--Langford and Catoni's bounds, revealing that the former is optimized by the Gibbs posterior. We introduce novel, tighter bounds for various types of loss functions. To achieve this, we introduce a new technique to optimize parameters in probabilistic statements. To study the limitations of these approaches, we present a counter-example where most of the information-theoretic bounds fail while traditional approaches do not. Finally, we explore the relationship between privacy and generalization. We show that algorithms with a bounded maximal leakage generalize. For discrete data, we derive new bounds for differentially private algorithms that guarantee generalization even with a constant privacy parameter, which is in contrast to previous bounds in the literature.

The realization of electron-spin resonance at the single-atom level using scanning tunneling microscopy has opened new avenues for coherent quantum sensing and quantum state manipulation at the ultimate size limit. This allows to build many-body Hamiltonians and the study of their complex physical behavior. Recently, a novel qubit platform has emerged from this field, raising questions about the driving mechanism from single-atom magnets. In this work, we demonstrate how single-atom magnets can be used to drive a nearby single spin qubit efficiently, while also addressing critical aspects related to the optimization of experimental parameters.

- Aims: We intended to quantify the impact of stellar multiplicity on the presence and properties of exoplanets. - Methods: We investigated all exoplanet host stars at less than 100 pc using the latest astrometric data from Gaia DR3 and advanced statistical methodologies. We complemented our search for common proper motion and parallax companions with data from the Washington Double Star catalogue and the literature. After excluding a number of systems based on radial velocity data, and membership in clusters and open associations, or with resolved ultracool companions, we kept 215 exoplanet host stars in 212 multiple-star systems. - Results: We found 17 new companions in the systems of 15 known exoplanet host stars, measured precise angular and projected physical separations and position angles for 236 pairs of stars, compiled key parameters for 276 planets in multiple systems, and established a comparison sample comprising 687 single stars with exoplanets. With all of this, we statistically analysed a series of hypothesis regarding planets in multiple stellar systems. Although they are only statistically significant at a 2{\sigma} level, our analysis pointed to several interesting results on the comparison in the mean number of planets in multiple versus single stellar systems and the tendency of high mass planets to be located in closer orbits in multiple systems. We confirm that planets in multiple systems tend to have orbits with larger eccentricities than those in single systems. In particular, we found a significant (> 4{\sigma}) preference for planets to exhibit high orbital eccentricities at small ratios between star-star projected physical separations and star-planet semi-major axes., Comment: Accepted in A&A. 47 Pages, 14 figures and 9 tables (4 tables online)

We report the discovery of five white dwarf + ultracool dwarf systems identified as common proper motion wide binaries in the Gaia Catalogue of Nearby Stars. The discoveries include a white dwarf + L subdwarf binary, VVV 1256-62AB, a gravitationally bound system located 75.6(+1.9/-1.8) pc away with a projected separation of 1375(+35/-33) au. The primary is a cool DC white dwarf with a hydrogen dominated atmosphere, and has a total age of 10.5(+3.3/-2.1) Gyr, based on white dwarf model fitting. The secondary is an L subdwarf with a metallicity of [M/H] = -0.72(+0.08/-0.10) (i.e. [Fe/H] = -0.81+/-0.10) and Teff = 2298(+45/-43) K based on atmospheric model fitting of its optical to near infrared spectrum, and likely has a mass just above the stellar/substellar boundary. The sub-solar metallicity of the L subdwarf and the system's total space velocity of 406 km/s indicates membership in the Galactic halo, and it has a flat eccentric Galactic orbit passing within 1~kpc of the centre of the Milky Way every ~0.4Gyr and extending to 15-31 kpc at apogal. VVV 1256-62B is the first L subdwarf to have a well-constrained age, making it an ideal benchmark of metal-poor ultracool dwarf atmospheres and evolution., Comment: 15 pages, 12 figures

Hyperspherical Prototypical Learning (HPL) is a supervised approach to representation learning that designs class prototypes on the unit hypersphere. The prototypes bias the representations to class separation in a scale invariant and known geometry. Previous approaches to HPL have either of the following shortcomings: (i) they follow an unprincipled optimisation procedure; or (ii) they are theoretically sound, but are constrained to only one possible latent dimension. In this paper, we address both shortcomings. To address (i), we present a principled optimisation procedure whose solution we show is optimal. To address (ii), we construct well-separated prototypes in a wide range of dimensions using linear block codes. Additionally, we give a full characterisation of the optimal prototype placement in terms of achievable and converse bounds, showing that our proposed methods are near-optimal., Comment: 14 pages: 9 of the main paper, 2 of references, and 3 of appendices. To appear in the Proceedings of the Geometry-grounded Representation Learning and Generative Modeling at the 41st International Conference on Machine Learning, Vienna, Austria. Code is available at: https://github.com/martinlindstrom/coding_theoretic_hpl

The stability and reproducibility issues related to metal halide Perovskite Solar Cells (PSCs) have prevented so far from exploiting this highly efficient photovoltaic technology under rainy or even moisture conditions and in combination with drop triboelectric nanogenerators (D-TENGs). Protecting the PSCs with a waterproof encapsulator could overcome such a disadvantage. Herein, we propose the implementation of fluorinated carbon (CFx) coatings produced by PECVD, enabling waterproof capability for the PSC and working as an energy harvesting surface for D-TENGs. These 100 nm - films present optimal optical transmittance, fully preserving the photovoltaic parameters of the cell. The improved long-term stability of the water-resistant PSCs prevents degradation under illumination in outdoor or simulated adverse environments with high humidity, high temperature, water immersion, or rain. As a remarkable result, the 50% Power Conversion Efficiency (PCE) has been retained after ten days of illumination under 100% relative humidity at 50 degrees Celsius. Also, the CFx coatings were successfully tested as a promoter agent to integrate commercial UV-curable sealants compatible with Spiro-OMeTAD, enhancing the performance stability of up to 80 % of PCE after 100 hours under illumination in a humid environment. This PSC was tested in a top-bottom electrode configuration for harvesting kinetic energy from droplets. Devices were compatible with simultaneously working as D-TENG and photovoltaic cells, yielding voltage outputs up to 12 V with maximum peak power density reaching 2.75 uW/cm2 defined by the D-TENG and PCE of 11.5 % and 8.46 mA/cm2 of short circuit current determined by the PSC under dripping and for an illumination angle of 45 degrees. The durability of the multisource device was tested under constant illumination and periodical drop impacting for more than 5 hours.

We present a computation of triply-heavy baryons on a quantum computer, employing the Cornell quark model in line with the earlier quarkonium work of Gallimore and Liao. These baryons are some of the most interesting Standard Model particles which have not yet been detected, as they bear on the short range (colour) behaviour of the nuclear force. The spectrum here obtained is compatible with predictions from earlier works, with our uncertainty dominated by traditional few-body approximations (size of the variational basis, center of mass recoil, parameter estimation...) and not by the statistical error from the quantum computer (deployed here as a small diagonalizer), which turns out to be negligible respect to the other sources of uncertainty, at least in the present unsophisticated few-body approximation. We have also substituted one or more heavy quarks for strange quarks., Comment: 16 pages, 7 plots, uses quantikz2

This paper introduces a highly efficient greedy decoding algorithm for Transducer-based speech recognition models. We redesign the standard nested-loop design for RNN-T decoding, swapping loops over frames and labels: the outer loop iterates over labels, while the inner loop iterates over frames searching for the next non-blank symbol. Additionally, we represent partial hypotheses in a special structure using CUDA tensors, supporting parallelized hypotheses manipulations. Experiments show that the label-looping algorithm is up to 2.0X faster than conventional batched decoding when using batch size 32. It can be further combined with other compiler or GPU call-related techniques to achieve even more speedup. Our algorithm is general-purpose and can work with both conventional Transducers and Token-and-Duration Transducers. We open-source our implementation to benefit the research community., Comment: Accepted at IEEE SLT 2024

The vast majority of inference time for RNN Transducer (RNN-T) models today is spent on decoding. Current state-of-the-art RNN-T decoding implementations leave the GPU idle ~80% of the time. Leveraging a new CUDA 12.4 feature, CUDA graph conditional nodes, we present an exact GPU-based implementation of greedy decoding for RNN-T models that eliminates this idle time. Our optimizations speed up a 1.1 billion parameter RNN-T model end-to-end by a factor of 2.5x. This technique can applied to the "label looping" alternative greedy decoding algorithm as well, achieving 1.7x and 1.4x end-to-end speedups when applied to 1.1 billion parameter RNN-T and Token and Duration Transducer models respectively. This work enables a 1.1 billion parameter RNN-T model to run only 16% slower than a similarly sized CTC model, contradicting the common belief that RNN-T models are not suitable for high throughput inference. The implementation is available in NVIDIA NeMo., Comment: Interspeech 2024 Proceedings

This paper is focused on the segregation of FGK dwarf and giant stars through narrow-band photometric data using the Spanish Virtual Observatory (SVO) Filter Profile Service and associated photometric tools. We selected spectra from the MILES, STELIB, and ELODIE stellar libraries, and used SVO photometric tools to derive the synthetic photometry in 15 J-PAS narrow filters, which were especially selected to cover spectral features sensitive to gravity changes. Using machine-learning techniques as the Gaussian mixture model and the support vector machine, we defined several criteria based on J-PAS colours to discriminate between dwarf and giant stars. We selected five colour-colour diagrams that presented the most promising separation between both samples. Our results show an overall accuracy in the studied sample of $\sim$0.97 for FGK stars, although a dependence on the luminosity type and the stellar effective temperature was found. We also defined a colour-temperature relation for dwarf stars with effective temperatures between 4\,000 and 7\,000\,K, which allows one to estimate the stellar effective temperature from four J-PAS filters ($J0450$, $J0510$, $J0550$, and $J0620$). Additionally, we extended the study to M-type giant and dwarf stars, achieving a similar accuracy to that for FGK stars., Comment: 19 pages, 18 figures, accepted for publication in A&A

We develop theoretical methods for the implementation of creation and destruction operators in separate registers of a quantum computer, allowing for a transparent and dynamical creation and destruction of particle modes in second quantization in problems with variable particle number. We establish theorems for the commutation (anticommutation) relations on a finite memory bank and provide the needed symmetrizing and antisymmetrizing operators. Finally, we provide formulae in terms of these operators for unitary evolution under conventional two- and four-body Hamiltonian terms, as well as terms varying the particle number. In this formalism, the number of qubits needed to codify $n$ particles with $N_p$ modes each is of order $n\log_2 N_p$. Such scaling is more efficient than the Jordan-Wigner transformation which requires $O(N_p)$ qubits, whenever there are a modest number of particles with a large number of states available to each (and less advantageous for a large number of particles with few states available to each). And although less efficient, it is also less cumbersome than compact encoding., Comment: 35 pages, 7 figures

The COVID-19 pandemic led to a large global effort to sequence SARS-CoV-2 genomes from patient samples to track viral evolution and inform public health response. Millions of SARS-CoV-2 genome sequences have been deposited in global public repositories. The Canadian COVID-19 Genomics Network (CanCOGeN - VirusSeq), a consortium tasked with coordinating expanded sequencing of SARS-CoV-2 genomes across Canada early in the pandemic, created the Canadian VirusSeq Data Portal, with associated data pipelines and procedures, to support these efforts. The goal of VirusSeq was to allow open access to Canadian SARS-CoV-2 genomic sequences and enhanced, standardized contextual data that were unavailable in other repositories and that meet FAIR standards (Findable, Accessible, Interoperable and Reusable). The Portal data submission pipeline contains data quality checking procedures and appropriate acknowledgement of data generators that encourages collaboration. Here we also highlight Duotang, a web platform that presents genomic epidemiology and modeling analyses on circulating and emerging SARS-CoV-2 variants in Canada. Duotang presents dynamic changes in variant composition of SARS-CoV-2 in Canada and by province, estimates variant growth, and displays complementary interactive visualizations, with a text overview of the current situation. The VirusSeq Data Portal and Duotang resources, alongside additional analyses and resources computed from the Portal (COVID-MVP, CoVizu), are all open-source and freely available. Together, they provide an updated picture of SARS-CoV-2 evolution to spur scientific discussions, inform public discourse, and support communication with and within public health authorities. They also serve as a framework for other jurisdictions interested in open, collaborative sequence data sharing and analyses.

The ultimate step towards the exploitation of water as a clean and renewable energy source addresses the energies stored in the low frequencies of liquid flows, which demands flexible solutions to adapt to multiple scenarios, from raindrops to waves, including water moving in pipelines and microdevices. Thus, harvesting low-frequency flows is a young concept compared to solar and wind powers, where triboelectric nanogenerators have been revealed as the most promising relevant actors. However, despite widespread attempts by researchers, the drop energy harvesters' output power is still low, mainly because of the limitations in candidates endowed with ideal triboelectric and wetting properties and also the non-optimal and centimetre-scale device architecture that prevents the conversion of the complete kinetic energy of impinging drops. Herein, we disclose a microscale triboelectric nanogenerator that can harvest a high density of electrical power from drops through a single, submillisecond, long-lasting step. The mechanism relies on an instantaneous electrical capacitance variation owing to the high-speed contact of the drops with the electrodes' active area. We discuss the role of the precharged effect of the triboelectric surface in the time characteristic of the conversion event. The capacitive and microscale structure of the device is endowed with a small form factor that allows for the production of densely packed arrays. The proposed architecture can be adjusted to different liquids and scales and is compatible with a variety of triboelectric surfaces, including flexible, transparent, and thin-film approaches., Comment: 23 pages, 6 figures in main text, 9 figures in Supporting Information, videos available from the authors

This paper studies the truncation method from Alquier [1] to derive high-probability PAC-Bayes bounds for unbounded losses with heavy tails. Assuming that the $p$-th moment is bounded, the resulting bounds interpolate between a slow rate $1 / \sqrt{n}$ when $p=2$, and a fast rate $1 / n$ when $p \to \infty$ and the loss is essentially bounded. Moreover, the paper derives a high-probability PAC-Bayes bound for losses with a bounded variance. This bound has an exponentially better dependence on the confidence parameter and the dependency measure than previous bounds in the literature. Finally, the paper extends all results to guarantees in expectation and single-draw PAC-Bayes. In order to so, it obtains analogues of the PAC-Bayes fast rate bound for bounded losses from [2] in these settings., Comment: 9 pages: 5 of main text, 1 of references, and 3 of appendices

This paper studies the Bayesian regret of a variant of the Thompson-Sampling algorithm for bandit problems. It builds upon the information-theoretic framework of [Russo and Van Roy, 2015] and, more specifically, on the rate-distortion analysis from [Dong and Van Roy, 2020], where they proved a bound with regret rate of $O(d\sqrt{T \log(T)})$ for the $d$-dimensional linear bandit setting. We focus on bandit problems with a metric action space and, using a chaining argument, we establish new bounds that depend on the metric entropy of the action space for a variant of Thompson-Sampling. Under suitable continuity assumption of the rewards, our bound offers a tight rate of $O(d\sqrt{T})$ for $d$-dimensional linear bandit problems., Comment: 15 pages: 8 of main text and 7 of appendices

Aims. Recent observational studies suggest that feedback from active galactic nuclei (AGNs) may play an important role in the formation and evolution of low-mass dwarf galaxies, an issue that has received little attention from a theoretical perspective. Methods. We investigate this using two sets of 12 cosmological magneto-hydrodynamical simulations of the formation of dwarf galaxies: one set using a version of the AURIGA galaxy formation physics model including AGN feedback and a parallel set with AGN feedback turned off. Results. We show that the full-physics AGN runs satisfactorily reproduce several scaling relations, including the M_ BH-M_*, M_ BH-sigma_* and the baryonic Tully-Fisher relation. We find that the global star formation (SF) of galaxies run with AGN is reduced compared to the one in which AGN has been turned off, suggesting that this type of feedback is a viable way of suppressing SF in low-mass dwarfs, even though none of our galaxies is completely quenched by z=0. Furthermore, we found a tight correlation between the median SF rates and the black-hole-to-stellar mass ratio (M_BH/M_star) in our simulated dwarfs. SF is suppressed due to gas heating in the vicinity of the AGN: less HI gas is available in AGN runs, though the total amount of gas is preserved across the two settings within each galaxy. This indicates that the main effect of AGN feedback in our dwarfs is to heat up and push the gas away from the galaxy's centre rather than expelling it completely. Finally, we show that the two galaxies harbouring the largest BHs have suffered a considerable (up to 65%) reduction in their central dark matter density, pinpointing the role of AGNs in determining the final dark matter mass distribution within dwarf galaxies. This pilot paper highlights the importance of modelling AGN feedback at the lowest mass scales and the impact this can have on dwarf galaxy evolution., Comment: 13 pages, 10 figures and 2 tables Submitted to A&A

The interference of single photons going through a double slit is a compelling demonstration of the wave and particle nature of light in the same experiment. Single photons produced by spontaneous parametric down-conversion can be used for this purpose. However, it is particularly challenging to implement due to coherency and resolution challenges. In this article, we present a tabletop laboratory arrangement suitable for the undergraduate instruction laboratory that overcomes these challenges. The apparatus scans a single detector to produce a plot showing the interference patterns of single photons. We include experimental data obtained using this setup demonstrating double-slit and single-slit interference as well as quantum erasing through the use of sheet polarizers., Comment: 22 pages, 6 figures, Accepted by American Journal of Physics