Search

This paper presents the design and technical progress of a precision X-Y stage for detector dithering and flexure compensation. The stage is being developed for use in the Magellan InfraRed Multi-Object Spectrograph, MIRMOS. MIRMOS is a very large Nasmyth mounted spectrograph containing a combination of refractive, reflective and diffractive optics mounted on a long cryogenic optical bench. The instrument utilizes five science cameras, each having a custom x-y stage to control the in-plane detector position within each camera, providing both dithering capability for improved sampling, and flexure compensation to correct for image motion that results from the gravity variant operation of the instrument. Designed to operate at 120~K, the stage will accurately control detector position in two orthogonal degrees of freedom, and have manual fine adjustment features to set detector tip, tilt and piston. The piezo-driven flexure stage provides high-resolution backlash-free motion of the detector and is very compact along the optical path, keeping camera length to a minimum. A magnetoresistive bridge provides position feedback in each degree of freedom, greatly reducing hysteresis, which is common in piezoelectric actuators. The system is designed to operate in open loop using a lookup table keyed to the Nasmyth rotator angle for flexure control. Here, the optomechanical design of the stage, electrical control system, and current performance results from early prototype efforts are presented and discussed., Comment: Proceedings of SPIE Astronomical Telescopes + Instrumentation 2024, 11 pages, 10 figures

The need for intelligent and efficient resource provisioning for the productive management of resources in real-world scenarios is growing with the evolution of telecommunications towards the 6G era. Technologies such as Open Radio Access Network (O-RAN) can help to build interoperable solutions for the management of complex systems. Probabilistic forecasting, in contrast to deterministic single-point estimators, can offer a different approach to resource allocation by quantifying the uncertainty of the generated predictions. This paper examines the cloud-native aspects of O-RAN together with the radio App (rApp) deployment options. The integration of probabilistic forecasting techniques as a rApp in O-RAN is also emphasized, along with case studies of real-world applications. Through a comparative analysis of forecasting models using the error metric, we show the advantages of Deep Autoregressive Recurrent network (DeepAR) over other deterministic probabilistic estimators. Furthermore, the simplicity of Simple-Feed-Forward (SFF) leads to a fast runtime but does not capture the temporal dependencies of the input data. Finally, we present some aspects related to the practical applicability of cloud-native O-RAN with probabilistic forecasting.

The dynamics of evolving fluid films in the viscous Stokes limit is relevant to various applications, such as the modeling of lipid bilayers in cells. While the governing equations were formulated by Scriven in 1960, solving for the flow of a deformable viscous surface with arbitrary shape and topology has remained a challenge. In this study, we present a straightforward discrete model based on variational principles to address this long-standing problem. We replace the classical equations, which are expressed with tensor calculus in local coordinate, with a simple coordinate-free, differential-geometric formulation. The formulation provides a fundamental understanding of the underlying mechanics and directly translates to discretization. We construct a discrete analogue of the system using the Onsager variational principle, which, in a smooth context, governs the flow of a viscous medium. In the discrete setting, instead of term-wise discretizing the coordinate-based Stokes equations, we construct a discrete Rayleighian for the system and derive the discrete Stokes equations via the variational principle. This approach results in a stable, structure-preserving variational integrator that solves the system on general manifolds.

Dissipative many-body quantum dynamics can feature strong symmetries which give rise to conserved quantities. We discuss here how a strong symmetry in conjunction with a nonequilibrium phase transition allows to devise a protocol for measuring collective many-body observables. To demonstrate this idea we consider a collective spin system whose constituents are governed by a dissipative dynamics that conserves the total angular momentum. We show that by continuously monitoring the system output the value of the total angular momentum can be inferred directly from the time-integrated emission signal, without the need of repeated projective measurements or reinitializations of the spins. This may offer a route towards the measurement of collective properties in qubit ensembles, with applications in quantum tomography, quantum computation and quantum metrology.

Effective policy and intervention strategies to combat human trafficking for child sexual exploitation material (CSEM) production require accurate prevalence estimates. Traditional Network Scale Up Method (NSUM) models often necessitate standalone surveys for each geographic region, escalating costs and complexity. This study introduces a partially pooled NSUM model, using a hierarchical Bayesian framework that efficiently aggregates and utilizes data across multiple regions without increasing sample sizes. We developed this model for a novel national survey dataset from the Philippines and we demonstrate its ability to produce detailed municipal-level prevalence estimates of trafficking for CSEM production. Our results not only underscore the model's precision in estimating hidden populations but also highlight its potential for broader application in other areas of social science and public health research, offering significant implications for resource allocation and intervention planning.

Multi-component metal nanoparticles (NPs) are of paramount importance in the chemical industry, as most processes therein employ heterogeneous catalysts. While these multi-component systems have been shown to result in higher product yields, improved selectivities, and greater stability through catalytic cycling, the structural dynamics of these materials in response to various stimuli (e.g. temperature, adsorbates, etc.) are not understood with atomistic resolution. Here, we present a highly accurate equivariant machine-learned force field (MLFF), constructed from ab initio training data collected using Bayesian active learning, that is able to reliably simulate PdAu surfaces and NPs in response to thermal treatment as well as exposure to reactive H$_2$ atmospheres. We thus provide a single model that is able to reliably describe the full space of geometric and chemical complexity for such a heterogeneous catalytic system across single crystals, gas-phase interactions, and NPs reacting with H$_2$, including catalyst degradation and explicit reactivity. Ultimately, we provide direct atomistic evidence that verifies existing experimental hypotheses for bimetallic catalyst deactivation under reaction conditions, namely that Pd preferentially segregates into the Au bulk through aggressive catalytic cycling and that this degradation is site-selective, as well as the reactivity for hydrogen exchange as a function of Pd ensemble size. We demonstrate that understanding of the atomistic evolution of these active sites is of the utmost importance, as it allows for design and control of material structure and corresponding performance, which can be vetted in silico., Comment: 19 pages, 8 figures (main + SI)

Sparse RGBD scene completion is a challenging task especially when considering consistent textures and geometries throughout the entire scene. Different from existing solutions that rely on human-designed text prompts or predefined camera trajectories, we propose GenRC, an automated training-free pipeline to complete a room-scale 3D mesh with high-fidelity textures. To achieve this, we first project the sparse RGBD images to a highly incomplete 3D mesh. Instead of iteratively generating novel views to fill in the void, we utilized our proposed E-Diffusion to generate a view-consistent panoramic RGBD image which ensures global geometry and appearance consistency. Furthermore, we maintain the input-output scene stylistic consistency through textual inversion to replace human-designed text prompts. To bridge the domain gap among datasets, E-Diffusion leverages models trained on large-scale datasets to generate diverse appearances. GenRC outperforms state-of-the-art methods under most appearance and geometric metrics on ScanNet and ARKitScenes datasets, even though GenRC is not trained on these datasets nor using predefined camera trajectories. Project page: https://minfenli.github.io/GenRC, Comment: ECCV 2024

The spectral features observed in kilonovae (KNe) reveal the elemental composition and the velocity structures of matter ejected from neutron star mergers. In the spectra of the kilonova AT2017gfo, a P Cygni line at about 1$\mu$m has been linked to Sr II, providing the first direct evidence of freshly synthesised $r$-process material. An alternative explanation to Sr II was proposed - He I $\lambda 1083.3$nm under certain non-local-thermodynamic-equilibrium (NLTE) conditions. A key way to robustly discriminate between these identifications, and indeed other proposed identifications, is to analyse the temporal emergence and evolution of the feature. In this analysis we trace the earliest appearance of the observed feature and detail its spectro-temporal evolution, which we compare with a collisional-radiative model of helium. We show that the 1$\mu$m P Cygni line is inconsistent with a He I interpretation both in emergence time and in subsequent spectral evolution. Self-consistent helium masses cannot reproduce the observed feature, due to the diminishing strength of radiative pathways leaving triplet helium., Comment: Submitted to A&A, 10 pages, 6 figures. Comments are welcome

Recent progress in neural rendering has brought forth pioneering methods, such as NeRF and Gaussian Splatting, which revolutionize view rendering across various domains like AR/VR, gaming, and content creation. While these methods excel at interpolating {\em within the training data}, the challenge of generalizing to new scenes and objects from very sparse views persists. Specifically, modeling 3D humans from sparse views presents formidable hurdles due to the inherent complexity of human geometry, resulting in inaccurate reconstructions of geometry and textures. To tackle this challenge, this paper leverages recent advancements in Gaussian Splatting and introduces a new method to learn generalizable human Gaussians that allows photorealistic and accurate view-rendering of a new human subject from a limited set of sparse views in a feed-forward manner. A pivotal innovation of our approach involves reformulating the learning of 3D Gaussian parameters into a regression process defined on the 2D UV space of a human template, which allows leveraging the strong geometry prior and the advantages of 2D convolutions. In addition, a multi-scaffold is proposed to effectively represent the offset details. Our method outperforms recent methods on both within-dataset generalization as well as cross-dataset generalization settings.

Laser powder bed fusion (LPBF) process can incur defects due to melt pool instabilities, spattering, temperature increase, and powder spread anomalies. Identifying defects through in-situ monitoring typically requires collecting, storing, and analyzing large amounts of data generated. The first goal of this work is to propose a new approach to accurately map in-situ data to a three-dimensional (3D) geometry, aiming to reduce the amount of storage. The second goal of this work is to introduce several new IR features for defect detection or process model calibration, which include laser scan order, local preheat temperature, maximum pre-laser scanning temperature, and number of spatters generated locally and their landing locations. For completeness, processing of other common IR features, such as interpass temperature, heat intensity, cooling rates, and melt pool area, are also presented with the underlying algorithm and Python implementation. A number of different parts are printed, monitored, and characterized to provide evidence of process defects and anomalies that different IR features are capable of detecting., Comment: 29 Pages, 19 Figures

Stochastic diffusion processes are pervasive in nature, from the seemingly erratic Brownian motion to the complex interactions of synaptically-coupled spiking neurons. Recently, drawing inspiration from Langevin dynamics, neuromorphic diffusion models were proposed and have become one of the major breakthroughs in the field of generative artificial intelligence. Unlike discriminative models that have been well developed to tackle classification or regression tasks, diffusion models as well as other generative models such as ChatGPT aim at creating content based upon contexts learned. However, the more complex algorithms of these models result in high computational costs using today's technologies, creating a bottleneck in their efficiency, and impeding further development. Here, we develop a spintronic voltage-controlled magnetoelectric memory hardware for the neuromorphic diffusion process. The in-memory computing capability of our spintronic devices goes beyond current Von Neumann architecture, where memory and computing units are separated. Together with the non-volatility of magnetic memory, we can achieve high-speed and low-cost computing, which is desirable for the increasing scale of generative models in the current era. We experimentally demonstrate that the hardware-based true random diffusion process can be implemented for image generation and achieve comparable image quality to software-based training as measured by the Frechet inception distance (FID) score, achieving ~10^3 better energy-per-bit-per-area over traditional hardware.

Femtosecond laser ablation is a process that bears both fundamental physics interest and has wide industrial applications. For decades, the lack of probes on the relevant time and length scales has prevented access to the highly nonequilibrium phase decomposition processes triggered by laser excitation. Enabled by the unprecedented intense femtosecond X-ray pulses delivered by an X-ray free electron laser, we report here results of time-resolved small angle scattering measurements on the dynamics of nanoscale phase decomposition in thin gold films upon femtosecond laser-induced ablation. By analyzing the features imprinted onto the small angle diffraction patterns, the transient heterogeneous density distributions within the ablation plume as obtained from molecular dynamics simulations get direct experimental confirmation., Comment: Main manuscript with 32 pages incl. 9 figures + supplementary materials with 16 pages incl. 5 figures

Various benchmarks have been proposed to test linguistic understanding in pre-trained vision \& language (VL) models. Here we build on the existence task from the VALSE benchmark (Parcalabescu et al, 2022) which we use to test models' understanding of negation, a particularly interesting issue for multimodal models. However, while such VL benchmarks are useful for measuring model performance, they do not reveal anything about the internal processes through which these models arrive at their outputs in such visio-linguistic tasks. We take inspiration from the growing literature on model interpretability to explain the behaviour of VL models on the understanding of negation. Specifically, we approach these questions through an in-depth analysis of the text encoder in CLIP (Radford et al, 2021), a highly influential VL model. We localise parts of the encoder that process negation and analyse the role of attention heads in this task. Our contributions are threefold. We demonstrate how methods from the language model interpretability literature (such as causal tracing) can be translated to multimodal models and tasks; we provide concrete insights into how CLIP processes negation on the VALSE existence task; and we highlight inherent limitations in the VALSE dataset as a benchmark for linguistic understanding., Comment: Accepted at the 3rd Workshop on Advances in Language and Vision Research (ALVR 2024)

Facial Action Unit (AU) detection has gained significant research attention as AUs contain complex expression information. In this paper, we unpack two fundamental factors in AU detection: data and subject identity regularization, respectively. Motivated by recent advances in foundation models, we highlight the importance of data and collect a diverse dataset Face9M, comprising 9 million facial images, from multiple public resources. Pretraining a masked autoencoder on Face9M yields strong performance in AU detection and facial expression tasks. We then show that subject identity in AU datasets provides a shortcut learning for the model and leads to sub-optimal solutions to AU predictions. To tackle this generic issue of AU tasks, we propose Identity Adversarial Training (IAT) and demonstrate that a strong IAT regularization is necessary to learn identity-invariant features. Furthermore, we elucidate the design space of IAT and empirically show that IAT circumvents the identity shortcut learning and results in a better solution. Our proposed methods, Facial Masked Autoencoder (FMAE) and IAT, are simple, generic and effective. Remarkably, the proposed FMAE-IAT approach achieves new state-of-the-art F1 scores on BP4D (67.1\%), BP4D+ (66.8\%), and DISFA (70.1\%) databases, significantly outperforming previous work. We release the code and model at https://github.com/forever208/FMAE-IAT, the first open-sourced facial model pretrained on 9 million diverse images.

The Large Binocular Telescope, with its expansive collecting area, angular resolving power, and advanced optical design, provides a robust platform for development and operation of advanced instrumentation for astronomical research. The LBT currently hosts a mature suite of instruments for spectroscopy and imaging at optical through mid-infrared wavelengths, supported by sophisticated adaptive optics systems. This contribution summarizes the current state of instrumentation, including upgrades to existing instruments and commissioning of second generation instruments now in progress. The LBT is soliciting proposals for next generation instrument concepts, with participation open to consortium members and others interested in participation in the Observatory., Comment: 6 pages, Proc. SPIE 13096-4, Ground-based and Airborne Instrumentation for Astronomy X, Yokohama 2024

Waveguide quantum electrodynamics, the study of atomic systems interacting with propagating electromagnetic fields, is a powerful platform for understanding the complex interplay between light and matter. Qubit control is an indispensable tool in this field, and most experiments have so far focused on narrowband electromagnetic waves that interact with qubits at specific frequencies. This interaction, however, changes significantly with fast, broadband pulses, as waveguide properties like dispersion affect the pulse evolution and its impact on the qubit. Here, we use dispersion to achieve spatial addressing of superconducting qubits separated by a sub-wavelength distance within a microwave waveguide. This novel approach relies on a self-focusing effect to create a position-dependent interaction between the pulse and the qubits. This experiment emphasizes the importance of dispersion in the design and analysis of quantum experiments, and offers new avenues for the rapid control of quantum states.

Generative pre-trained transformer (GPT) models have shown promise in clinical entity and relation extraction tasks because of their precise extraction and contextual understanding capability. In this work, we further leverage the Unified Medical Language System (UMLS) knowledge base to accurately identify medical concepts and improve clinical entity and relation extraction at the document level. Our framework selects UMLS concepts relevant to the text and combines them with prompts to guide language models in extracting entities. Our experiments demonstrate that this initial concept mapping and the inclusion of these mapped concepts in the prompts improves extraction results compared to few-shot extraction tasks on generic language models that do not leverage UMLS. Further, our results show that this approach is more effective than the standard Retrieval Augmented Generation (RAG) technique, where retrieved data is compared with prompt embeddings to generate results. Overall, we find that integrating UMLS concepts with GPT models significantly improves entity and relation identification, outperforming the baseline and RAG models. By combining the precise concept mapping capability of knowledge-based approaches like UMLS with the contextual understanding capability of GPT, our method highlights the potential of these approaches in specialized domains like healthcare., Comment: Accepted at Association for Computational Linguistics BioNLP 2024

Recent research has demonstrated the feasibility of training efficient intent detectors based on pre-trained language model~(PLM) with limited labeled data. However, deploying these detectors in resource-constrained environments such as mobile devices poses challenges due to their large sizes. In this work, we aim to address this issue by exploring techniques to minimize the size of PLM-based intent detectors trained with few-shot data. Specifically, we utilize large language models (LLMs) for data augmentation, employ a cutting-edge model compression method for knowledge distillation, and devise a vocabulary pruning mechanism called V-Prune. Through these approaches, we successfully achieve a compression ratio of 21 in model memory usage, including both Transformer and the vocabulary, while maintaining almost identical performance levels on four real-world benchmarks.

A wide array of sequence models are built on a framework modeled after Transformers, comprising alternating sequence mixer and channel mixer layers. This paper studies a unifying matrix mixer view of sequence mixers that can be conceptualized as a linear map on the input sequence. This framework encompasses a broad range of well-known sequence models, including the self-attention of Transformers as well as recent strong alternatives such as structured state space models (SSMs), and allows understanding downstream characteristics such as efficiency and expressivity through properties of their structured matrix class. We identify a key axis of matrix parameterizations termed sequence alignment, which increases the flexibility and performance of matrix mixers, providing insights into the strong performance of Transformers and recent SSMs such as Mamba. Furthermore, the matrix mixer framework offers a systematic approach to developing sequence mixers with desired properties, allowing us to develop several new sub-quadratic sequence models. In particular, we propose a natural bidirectional extension of the Mamba model (Hydra), parameterized as a quasiseparable matrix mixer, which demonstrates superior performance over other sequence models including Transformers on non-causal tasks. As a drop-in replacement for attention layers, Hydra outperforms BERT by 0.8 points on the GLUE benchmark and ViT by 2% Top-1 accuracy on ImageNet.

Accurate estimate of long-term risk is critical for safe decision-making, but sampling from rare risk events and long-term trajectories can be prohibitively costly. Risk gradient can be used in many first-order techniques for learning and control methods, but gradient estimate is difficult to obtain using Monte Carlo (MC) methods because the infinitesimal divisor may significantly amplify sampling noise. Motivated by this gap, we propose an efficient method to evaluate long-term risk probabilities and their gradients using short-term samples without sufficient risk events. We first derive that four types of long-term risk probability are solutions of certain partial differential equations (PDEs). Then, we propose a physics-informed learning technique that integrates data and physics information (aforementioned PDEs). The physics information helps propagate information beyond available data and obtain provable generalization beyond available data, which in turn enables long-term risk to be estimated using short-term samples of safe events. Finally, we demonstrate in simulation that the proposed technique has improved sample efficiency, generalizes well to unseen regions, and adapts to changing system parameters., Comment: arXiv admin note: substantial text overlap with arXiv:2305.06432

HESS J1809-193 is an unidentified TeV source, first detected by the High Energy Stereoscopic System (H.E.S.S.) Collaboration. The emission originates in a source-rich region that includes several Supernova Remnants (SNR) and Pulsars (PSR) including SNR G11.1+0.1, SNR G11.0-0.0, and the young radio pulsar J1809-1917. Originally classified as a pulsar wind nebula (PWN) candidate, recent studies show the peak of the TeV region overlapping with a system of molecular clouds. This resulted in the revision of the original leptonic scenario to look for alternate hadronic scenarios. Marked as a potential PeVatron candidate, this region has been studied extensively by H.E.S.S. due to its emission extending up-to several tens of TeV. In this work, we use 2398 days of data from the High Altitude Water Cherenkov (HAWC) observatory to carry out a systematic source search for the HESS J1809-193 region. We were able to resolve emission detected as an extended component (modelled as a Symmetric Gaussian with a 1 $\sigma$ radius of 0.21 $^\circ$) with no clear cutoff at high energies and emitting photons up-to 210 TeV. We model the multi-wavelength observations for the region HESS J1809-193 using a time-dependent leptonic model and a lepto-hadronic model. Our model indicates that both scenarios could explain the observed data within the region of HESS J1809-193.

Coincident multi-messenger observation of cosmic sources can offer numerous benefits. One significant advantage is enhancing the detection significance of separate detectors by correlating their data and assuming a joint emission. We have formulated an approach for updating the Bayesian posterior probability of an astrophysical origin, namely $p_{\rm astro}$, relying on multi-messenger coincidences. We demonstrated this with candidate coincident gravitational waves and high-energy neutrinos. Applying our method to the public data of candidate coincident high-energy neutrinos with subthreshold gravitational-wave triggers, we found that in the case of highly energetic neutrino coincidences, $p_{\rm astro}$ can increase from approximately $\sim 0.1$ to $\sim 0.9$. This marked improvement makes subthreshold detections much more confident., Comment: 7 pages, 1 figure, 1 table

We prove the local Lipschitz regularity for solutions to an orthotropic $q$ Laplacian-type equation in the Heisenberg group. The technique we apply is largely inspired by the works of X. Zhong, where the $q$-Laplacian was studied in the Heinsenberg group and the H\"older regularity of gradient was proved. Due to the degeneracy of the present equation, this regularity of the gradient is not even known in the Euclidean setting, and only its boundedness is expected.

Functional properties of mixed ionic electronic conductors (MIECs) can be radically modified by (de)insertion of mobile charged defects. A complete control of this dynamic behaviour has multiple applications in a myriad of fields including advanced computing, data processing, sensing or energy conversion. However, the effect of different MIECs state-of-charge is not fully understood yet and there is a lack of strategies for fully controlling the defect content in a material. In this work we present a model-less technique to characterize ionic defect concentration and ionic insertion kinetics in MIEC materials: Optoionic Impedance Spectroscopy (OIS). The proof of concept and advantages of OIS are demonstrated by studying the oxygen (de)insertion in thin films of hole-doped perovskite oxides. Ion migration into/out of the studied materials is achieved by the application of an electrochemical potential, achieving stable and reversible modification of its optical properties. By tracking the dynamic variation of optical properties depending on the gating conditions, OIS enables to extract electrochemical parameters involved in the electrochromic process. The results demonstrate the capability of the technique to effectively characterize the kinetics of single- and even multi- layer systems. The technique can be employed for studying underlying mechanisms of the response characteristics of MIEC-based devices.

Based on the rate of change of its orbital period, PSR J2043+1711 has a substantial peculiar acceleration of 3.5 $\pm$ 0.8 mm/s/yr, which deviates from the acceleration predicted by equilibrium Milky Way models at a $4\sigma$ level. The magnitude of the peculiar acceleration is too large to be explained by disequilibrium effects of the Milky Way interacting with orbiting dwarf galaxies ($\sim$1 mm/s/yr), and too small to be caused by period variations due to the pulsar being a redback. We identify and examine two plausible causes for the anomalous acceleration: a stellar flyby, and a long-period orbital companion. We identify a main-sequence star in \textit{Gaia} DR3 and Pan-STARRS DR2 with the correct mass, distance, and on-sky position to potentially explain the observed peculiar acceleration. However, the star and the pulsar system have substantially different proper motions, indicating that they are not gravitationally bound. However, it is possible that this is an unrelated star that just happens to be located near J2043+1711 along our line of sight (chance probability of 1.6\%). Therefore, we also constrain possible orbital parameters for a circumbinary companion in a hierarchical triple system with J2043+1711; the changes in the spindown rate of the pulsar are consistent with an outer object that has an orbital period of 80 kyr, a companion mass of 0.3 $M_\odot$ (indicative of a white dwarf or low-mass star), and a semi-major axis of 2000 AU. Continued timing and/or future faint optical observations of J2043+1711 may eventually allow us to differentiate between these scenarios.

Galaxy clusters are the most massive, gravitationally-bound structures in the Universe, emerging through hierarchical structure formation of large-scale dark matter and baryon overdensities. Early galaxy ``proto-clusters'' are believed to be important physical drivers of the overall cosmic star-formation rate density and serve as ``hotspots'' for the reionization of the intergalactic medium. Our understanding of the formation of these structures at the earliest cosmic epochs is, however, limited to sparse observations of their galaxy members, or based on phenomenological models and cosmological simulations. Here we report the detection of a massive neutral, atomic hydrogen (HI) gas reservoir permeating a galaxy proto-cluster at redshift $z=5.4$, observed one billion years after the Big Bang. The presence of this cold gas is revealed by strong damped Lyman-$\alpha$ absorption features observed in several background galaxy spectra taken with JWST/NIRSpec in close on-sky projection. While overall the sightlines probe a large range in HI column densities, $N_{\rm HI} = 10^{21.7}-10^{23.5}$ cm$^{-2}$, they are similar across nearby sightlines, demonstrating that they probe the same dense, neutral gas. This observation of a massive, large-scale overdensity of cold neutral gas challenges current large-scale cosmological simulations and has strong implications for the reionization topology of the Universe., Comment: Submitted

Training a classifier on web-crawled data demands learning algorithms that are robust to annotation errors and irrelevant examples. This paper builds upon the recent empirical observation that applying unsupervised contrastive learning to noisy, web-crawled datasets yields a feature representation under which the in-distribution (ID) and out-of-distribution (OOD) samples are linearly separable. We show that direct estimation of the separating hyperplane can indeed offer an accurate detection of OOD samples, and yet, surprisingly, this detection does not translate into gains in classification accuracy. Digging deeper into this phenomenon, we discover that the near-perfect detection misses a type of clean examples that are valuable for supervised learning. These examples often represent visually simple images, which are relatively easy to identify as clean examples using standard loss- or distance-based methods despite being poorly separated from the OOD distribution using unsupervised learning. Because we further observe a low correlation with SOTA metrics, this urges us to propose a hybrid solution that alternates between noise detection using linear separation and a state-of-the-art (SOTA) small-loss approach. When combined with the SOTA algorithm PLS, we substantially improve SOTA results for real-world image classification in the presence of web noise github.com/PaulAlbert31/LSA, Comment: Accepted in the European Conference on Computer Vision (ECCV) 2024

Combining the predictions of multiple trained models through ensembling is generally a good way to improve accuracy by leveraging the different learned features of the models, however it comes with high computational and storage costs. Model fusion, the act of merging multiple models into one by combining their parameters reduces these costs but doesn't work as well in practice. Indeed, neural network loss landscapes are high-dimensional and non-convex and the minima found through learning are typically separated by high loss barriers. Numerous recent works have been focused on finding permutations matching one network features to the features of a second one, lowering the loss barrier on the linear path between them in parameter space. However, permutations are restrictive since they assume a one-to-one mapping between the different models' neurons exists. We propose a new model merging algorithm, CCA Merge, which is based on Canonical Correlation Analysis and aims to maximize the correlations between linear combinations of the model features. We show that our alignment method leads to better performances than past methods when averaging models trained on the same, or differing data splits. We also extend this analysis into the harder setting where more than 2 models are merged, and we find that CCA Merge works significantly better than past methods. Our code is publicly available at https://github.com/shoroi/align-n-merge, Comment: Proceedings of the Forty-first International Conference on Machine Learning (ICML 2024)

The fermionic R\'enyi entanglement entropy is studied for causal diamonds in two-dimensional Minkowski spacetime. Choosing the quasi-free state describing the Minkowski vacuum with an ultraviolet regularization, a logarithmically enhanced area law is derived., Comment: 15 pages, LaTeX, 1 figure

Broadband absorption in the visible spectrum is essential in optoelectronic applications that involve power conversion such as photovoltaics and photocatalysis. Most ultrathin broadband absorbers use parasitic plasmonic structures that maximize absorption using surface plasmons and/or Fabry-Perot cavities, which limits the weight efficiency of the device. Here, we show the theoretical and experimental realization of an unpatterned/planar semiconductor thin-film absorber based on monolayer transition metal dichalcogenides (TMDCs). We experimentally demonstrate an average total absorption in the visible range (450 nm - 700 nm) of > 70% using > 4 nm of semiconductor absorbing materials scalable over large areas with vapor phase growth techniques. Our analysis suggests that a power conversion efficiency (PCE) of 15.54% and a specific power > 300 W g^-1 may be achieved in a photovoltaic cell based on this metamaterial absorber.

Blockchain databases have attracted widespread attention but suffer from poor scalability due to underlying non-scalable blockchains. While blockchain sharding is necessary for a scalable blockchain database, it poses a new challenge named on-chain cross-shard database services. Each cross-shard database service (e.g., cross-shard queries or inter-shard load balancing) involves massive cross-shard data exchanges, while the existing cross-shard mechanisms need to process each cross-shard data exchange via the consensus of all nodes in the related shards (i.e., on-chain) to resist a Byzantine environment of blockchain, which eliminates sharding benefits. To tackle the challenge, this paper presents GriDB, the first scalable blockchain database, by designing a novel off-chain cross-shard mechanism for efficient cross-shard database services. Borrowing the idea of off-chain payments, GriDB delegates massive cross-shard data exchange to a few nodes, each of which is randomly picked from a different shard. Considering the Byzantine environment, the untrusted delegates cooperate to generate succinct proof for cross-shard data exchanges, while the consensus is only responsible for the low-cost proof verification. However, different from payments, the database services' verification has more requirements (e.g., completeness, correctness, freshness, and availability); thus, we introduce several new authenticated data structures (ADS). Particularly, we utilize consensus to extend the threat model and reduce the complexity of traditional accumulator-based ADS for verifiable cross-shard queries with a rich set of relational operators. Moreover, we study the necessity of inter-shard load balancing for a scalable blockchain database and design an off-chain and live approach for both efficiency and availability during balancing.

We report an observation of ultra-high energy (UHE) gamma rays from the Galactic Center region, using seven years of data collected by the High-Altitude Water Cherenkov (HAWC) Observatory. The HAWC data are best described as a point-like source (HAWC J1746-2856) with a power-law spectrum ($\mathrm{d}N/\mathrm{d}E=\phi(E/26 \,\text{TeV})^{\gamma}$), where $\gamma=-2.88 \pm 0.15_{\text{stat}} - 0.1_{\text{sys}} $ and $\phi=1.5 \times 10^{-15}$ (TeV cm$^{2}$s)$^{-1}$ $\pm\, 0.3_{\text{stat}}\,^{+0.08_{\text{sys}}}_{-0.13_{\text{sys}}}$ extending from 6 to 114 TeV. We find no evidence of a spectral cutoff up to $100$ TeV using HAWC data. Two known point-like gamma-ray sources are spatially coincident with the HAWC gamma-ray excess: Sgr A$^{*}$ (HESS J1745-290) and the Arc (HESS J1746-285). We subtract the known flux contribution of these point sources from the measured flux of HAWC J1746-2856 to exclude their contamination and show that the excess observed by HAWC remains significant ($>$5$\sigma$) with the spectrum extending to $>$100 TeV. Our result supports that these detected UHE gamma rays can originate via hadronic interaction of PeV cosmic-ray protons with the dense ambient gas and confirms the presence of a proton PeVatron at the Galactic Center.

Pre-trained models produce strong generic representations that can be adapted via fine-tuning. The learned weight difference relative to the pre-trained model, known as a task vector, characterises the direction and stride of fine-tuning. The significance of task vectors is such that simple arithmetic operations on them can be used to combine diverse representations from different domains. This paper builds on these properties of task vectors and aims to answer (1) whether components of task vectors, particularly parameter blocks, exhibit similar characteristics, and (2) how such blocks can be used to enhance knowledge composition and transfer. To this end, we introduce aTLAS, an algorithm that linearly combines parameter blocks with different learned coefficients, resulting in anisotropic scaling at the task vector level. We show that such linear combinations explicitly exploit the low intrinsic dimensionality of pre-trained models, with only a few coefficients being the learnable parameters. Furthermore, composition of parameter blocks leverages the already learned representations, thereby reducing the dependency on large amounts of data. We demonstrate the effectiveness of our method in task arithmetic, few-shot recognition and test-time adaptation, with supervised or unsupervised objectives. In particular, we show that (1) learned anisotropic scaling allows task vectors to be more disentangled, causing less interference in composition; (2) task vector composition excels with scarce or no labeled data and is less prone to domain shift, thus leading to better generalisability; (3) mixing the most informative parameter blocks across different task vectors prior to training can reduce the memory footprint and improve the flexibility of knowledge transfer. Moreover, we show the potential of aTLAS as a PEFT method, particularly with less data, and demonstrate that its scalibility.

Environmental Insights Explorer (EIE) is a Google product that reports aggregate statistics about human mobility, including various methods of transit used by people across roughly 50,000 regions globally. These statistics are used to estimate carbon emissions and provided to policymakers to inform their decisions on transportation policy and infrastructure. Due to the inherent sensitivity of this type of user data, it is crucial that the statistics derived and released from it are computed with appropriate privacy protections. In this work, we use a combination of federated analytics and differential privacy to release these required statistics, while operating under strict error constraints to ensure utility for downstream stakeholders. In this work, we propose a new mechanism that achieves $ \epsilon \approx 2 $-DP while satisfying these strict utility constraints, greatly improving over natural baselines. We believe this mechanism may be of more general interest for the broad class of group-by-sum workloads.

We provide a unified framework to certify nonlocal properties of quantum channels from the correlations obtained in measurement protocols. Our approach gathers and extends fully- and semi-device independent methods for this purpose. We study the effect of different models of dephasing noise, some of which are shown to generate nonlocality and entanglement in special cases. In the extreme case of complete dephasing, the measurement protocols discussed yield particularly simple tests to certify nonlocality. These are based on the relations between bipartite quantum channels and their classical analogs: bipartite stochastic matrices defining conditional distributions.

Biobanks advance biomedical and clinical research by collecting and offering data and biological samples for numerous studies. However, the impact of these repositories varies greatly due to differences in their purpose, scope, governance, and data collected. Here, we computationally identified 2,663 biobanks and their textual mentions in 228,761 scientific articles, 16,210 grants, 15,469 patents, 1,769 clinical trials, and 9,468 public policy documents, helping characterize the academic communities that utilize and support them. We found a strong concentration of biobank-related research on a few diseases, where 20\% of publications focus on obesity, Alzheimer's disease, breast cancer, and diabetes. Moreover, collaboration, rather than citation count, shapes the community's recognition of a biobank. We show that, on average, 41.1\% of articles miss to reference any of the biobank's reference papers and 59.6\% include a biobank member as a co-author. Using a generalized linear model, we identified the key factors that contribute to the impact of a biobank, finding that an impactful biobank tends to be more open to external researchers, and that quality data -- especially linked medical records -- as opposed to large data, correlates with a higher impact in science, innovation, and disease. The collected data and findings are accessible through an open-access web application intended to inform strategies to expand access and maximize the value of these valuable resources., Comment: 14 pages, 5 figures

One of the more complex tasks for researchers using HPC systems is performance monitoring and tuning of their applications. Developing a practice of continuous performance improvement, both for speed-up and efficient use of resources is essential to the long term success of both the HPC practitioner and the research project. Profiling tools provide a nice view of the performance of an application but often have a steep learning curve and rarely provide an easy to interpret view of resource utilization. Lower level tools such as top and htop provide a view of resource utilization for those familiar and comfortable with Linux but a barrier for newer HPC practitioners. To expand the existing profiling and job monitoring options, the MIT Lincoln Laboratory Supercomputing Center created LLoad, a tool that captures a snapshot of the resources being used by a job on a per user basis. LLload is a tool built from standard HPC tools that provides an easy way for a researcher to track resource usage of active jobs. We explain how the tool was designed and implemented and provide insight into how it is used to aid new researchers in developing their performance monitoring skills as well as guide researchers in their resource requests.

We consider the well-posedness of the family of dispersion generalized Benjamin-Ono equations. Earlier work of Herr-Ionescu-Kenig-Koch established well-posedness with data in $L^2$, by using a discretized gauge transform in the setting of Bourgain spaces. In this article, we remain in the simpler functional setting of Sobolev spaces, and instead combine a pseudodifferential gauge transform, a paradifferential normal form, and a variable coefficient Strichartz analysis to establish well-posedness in negative-exponent Sobolev spaces. Our result coincides with the classical well-posedness results obtained at the Benjamin-Ono and KdV endpoints., Comment: 40 pages

High-significance evidences of the existence of a high-energy diffuse flux of cosmic neutrinos have emerged in the last decade from several observations by the IceCube Collaboration. The ANTARES neutrino telescope took data for 15 years in the Mediterranean Sea, from 2007 to 2022, and collected a high-purity all-flavour neutrino sample. The search for a diffuse cosmic neutrino signal using this dataset is presented in this article. This final analysis did not provide a statistically significant observation of the cosmic diffuse flux: this is converted into limits on the properties of the cosmic neutrino spectrum. In particular, given the sensitivity of the ANTARES neutrino telescope between 1 and 50 TeV, constraints on single-power-law hypotheses are derived for the cosmic diffuse flux below 20 TeV.

Multi-component quantum mixtures in one dimension can be characterized by their symmetry under particle exchange. For a strongly-interacting Bose-Bose mixture, we show that the time evolution of the momentum distribution from an initially symmetry-mixed state is quasi-constant for a SU(2) symmetry conserving Hamiltonian, while it displays large oscillations in time for the symmetry-breaking case where inter- and intra-species interactions are different. Using the property that the momentum distribution operator at strong interactions commutes with the class-sum operator, the latter acting as a symmetry witness, we show that the momentum distribution oscillations correspond to symmetry oscillations, with a mechanism analog to neutrino flavour oscillations.

Super-Earths orbiting M-dwarf stars may be the most common habitable planets in the Universe. However, their habitability is threatened by intense irradiation from their host stars, which drives the escape of water to space and can lead to surface desiccation. We present simulation results of a box model of water cycling between interior and atmosphere and loss to space, for terrestrial planets of mass 1--8 $M_\oplus$ orbiting in the habitable zone of a late M-dwarf. Energy-limited loss decreases with planetary mass, while diffusion-limited loss increases with mass. Depending on where it orbits in the habitable zone, a 1 $M_\oplus$ planet that starts with 3--8 Earth Oceans can end up with an Earth-like surface of oceans and exposed continents; for an 8 $M_\oplus$ super-Earth, that range is 3--12 Earth Oceans. Planets initialized with more water end up as waterworlds with no exposed continents, while planets that start with less water have desiccated surfaces by 5 Gyr. Since the mantles of terrestrial planets can hold much more water than is currently present in Earth's atmosphere, none of our simulations result in Dune planets -- such planets may be less common than previously thought. Further, more water becomes sequestered within the mantle for larger planets. A super-Earth at the inner edge of the habitable zone tends to end up as either a waterworld or with a desiccated surface; only a narrow range of initial water inventory yields an Earth-like surface., Comment: 12 pages, 5 figures; revisions submitted to The Astrophysical Journal

Video understanding requires the extraction of rich spatio-temporal representations, which transformer models achieve through self-attention. Unfortunately, self-attention poses a computational burden. In NLP, Mamba has surfaced as an efficient alternative for transformers. However, Mamba's successes do not trivially extend to computer vision tasks, including those in video analysis. In this paper, we theoretically analyze the differences between self-attention and Mamba. We identify two limitations in Mamba's token processing: historical decay and element contradiction. We propose VideoMambaPro (VMP) that solves the identified limitations by adding masked backward computation and elemental residual connections to a VideoMamba backbone. VideoMambaPro shows state-of-the-art video action recognition performance compared to transformer models, and surpasses VideoMamba by clear margins: 7.9% and 8.1% top-1 on Kinetics-400 and Something-Something V2, respectively. Our VideoMambaPro-M model achieves 91.9% top-1 on Kinetics-400, only 0.2% below InternVideo2-6B but with only 1.2% of its parameters. The combination of high performance and efficiency makes VideoMambaPro an interesting alternative for transformer models.

Transformers are extremely successful machine learning models whose mathematical properties remain poorly understood. Here, we rigorously characterize the behavior of transformers with hardmax self-attention and normalization sublayers as the number of layers tends to infinity. By viewing such transformers as discrete-time dynamical systems describing the evolution of points in a Euclidean space, and thanks to a geometric interpretation of the self-attention mechanism based on hyperplane separation, we show that the transformer inputs asymptotically converge to a clustered equilibrium determined by special points called leaders. We then leverage this theoretical understanding to solve sentiment analysis problems from language processing using a fully interpretable transformer model, which effectively captures `context' by clustering meaningless words around leader words carrying the most meaning. Finally, we outline remaining challenges to bridge the gap between the mathematical analysis of transformers and their real-life implementation., Comment: 23 pages, 10 figures, 1 table. Funded by the European Union (Horizon Europe MSCA project ModConFlex, grant number 101073558). Accompanying code available at: https://github.com/DCN-FAU-AvH/clustering-hardmax-transformers

We introduce a modal logic FIL for Feferman interpretability. In this logic both the provability modality and the interpretability modality can come with a label. This label indicates that in the arithmetical interpretation the axiom set of the underlying base theory is tweaked so as to mimic behaviour of finitely axiomatised theories. The theory with the tweaked axiom set will be extensionally the same as the original theory though this equality will in general not be provable. After providing the logic FIL and proving the arithmetical soundness, we set the logic to work to prove various interpretability principles to be sound in a large variety of (weak) arithmetical theories. In particular, we prove the two series of principles from [GJ20] to be arithmetically sound using FIL. Up to date, the arithmetical soundness of these series had only been proven using the techniques of definable cuts.

There is an increasing trend towards evaluating NLP models with LLM-generated judgments instead of human judgments. In the absence of a comparison against human data, this raises concerns about the validity of these evaluations; in case they are conducted with proprietary models, this also raises concerns over reproducibility. We provide JUDGE-BENCH, a collection of 20 NLP datasets with human annotations, and comprehensively evaluate 11 current LLMs, covering both open-weight and proprietary models, for their ability to replicate the annotations. Our evaluations show that each LLM exhibits a large variance across datasets in its correlation to human judgments. We conclude that LLMs are not yet ready to systematically replace human judges in NLP.

Non-linear transport effects in response to external magnetic fields, i.e. electrical magnetochiral anisotropy (eMChA), have attracted much attention for their importance to study quantum and spin-related phenomena. Indeed, they have permitted the exploration of topological surface states and charge-to-spin conversion processes in low-symmetry systems. Nevertheless, despite the inherent correlation between the symmetry of the material under examination and its non-linear transport characteristics, there is a lack of experimental demonstration to delve into this relationship and to unveil their microscopic mechanisms. Here, we study eMChA in chiral elemental Tellurium (Te) along different crystallographic directions, establishing the connection between the different eMChA components and the crystal symmetry of Te. We observed different longitudinal eMChA components with collinear current and magnetic field, demonstrating experimentally the radial angular momentum texture of Te. We also measured a transverse non-linear resistance which, as the longitudinal counterpart, scales bilinearly with current and magnetic fields, illustrating that they are different manifestations of the same effect. Finally, we study the scaling law of the eMChA, evidencing that extrinsic scattering from dynamic sources is the dominant microscopic mechanism. These findings underscore the efficacy of symmetry-based investigations in understanding and predicting non-linear transport phenomena, with potential applications in spintronics and energy harvesting., Comment: 8 pages, 4 figures, and Supplemental Material

In this study, we investigate the impact of a periodic array of hydrophilic cylindrical obstacles on the spatio-temporal dynamics of chemical waves in the bubble-free Belousov-Zhabotinsky-1,4-cyclohexanedione (BZ-CHD) oscillatory chemical reaction. We find that the thin fluid layer ascending the outer surfaces of these hydrophilic obstacles is crucial for influencing wave dynamics. We observe that circular waves begin almost simultaneously at these obstacles and spread outward. In a covered setup where evaporation is minimal, these wavefronts retain their circular shape. However, in an open setup where evaporative cooling is more pronounced, thermal Marangoni flows trigger wavefront instabilities, leading to the formation of distinctive flower-like patterns around the obstacles. Our experiments also shows that the number of petals formed increases linearly with the diameter of the pillars, though a minimum diameter is required for the instabilities to manifest. These results highlight the ability to engineer specific wave patterns, offering a strategy to control and steer the reaction dynamics. This capability is particularly crucial for the development of microfluidic devices that demand precise control over chemical wave propagation.

This paper presents the design, calibration, and survey strategy of the Fast Radio Burst (FRB) digital backend and its real-time data processing pipeline employed in the Tianlai Cylinder Pathfinder array. The array, consisting of three parallel cylindrical reflectors and equipped with 96 dual-polarization feeds, is a radio interferometer array designed for conducting drift scans of the northern celestial semi-sphere. The FRB digital backend enables the formation of 96 digital beams, effectively covering an area of approximately 40 square degrees with 3 dB beam. Our pipeline demonstrates the capability to make automatic search of FRBs, detecting at quasi-real-time and classify FRB candidates automatically. The current FRB searching pipeline has an overall recall rate of 88\%. During the commissioning phase, we successfully detected signals emitted by four well-known pulsars: PSR B0329+54, B2021+51, B0823+26, and B2020+28. We report the first discovery of an FRB by our array, designated as FRB 20220414A. We also investigate the optimal arrangement for the digitally formed beams to achieve maximum detection rate by numerical simulation., Comment: 27 pages, 21 figures, 7 tables, RAA accepted

LHS 1140 b is the second-closest temperate transiting planet to the Earth with an equilibrium temperature low enough to support surface liquid water. At 1.730$\pm$0.025 R$_\oplus$, LHS 1140 b falls within the radius valley separating H$_2$-rich mini-Neptunes from rocky super-Earths. Recent mass and radius revisions indicate a bulk density significantly lower than expected for an Earth-like rocky interior, suggesting that LHS 1140 b could either be a mini-Neptune with a small envelope of hydrogen ($\sim$0.1% by mass) or a water world (9--19% water by mass). Atmospheric characterization through transmission spectroscopy can readily discern between these two scenarios. Here, we present two JWST/NIRISS transit observations of LHS 1140 b, one of which captures a serendipitous transit of LHS 1140 c. The combined transmission spectrum of LHS 1140 b shows a telltale spectral signature of unocculted faculae (5.8 $\sigma$), covering $\sim$20% of the visible stellar surface. Besides faculae, our spectral retrieval analysis reveals tentative evidence of residual spectral features, best-fit by Rayleigh scattering from an N$_2$-dominated atmosphere (2.3 $\sigma$), irrespective of the consideration of atmospheric hazes. We also show through Global Climate Models (GCM) that H$_2$-rich atmospheres of various compositions (100$\times$, 300$\times$, 1000$\times$solar metallicity) are ruled out to $>$10 $\sigma$. The GCM calculations predict that water clouds form below the transit photosphere, limiting their impact on transmission data. Our observations suggest that LHS 1140 b is either airless or, more likely, surrounded by an atmosphere with a high mean molecular weight. Our tentative evidence of an N$_2$-rich atmosphere provides strong motivation for future transmission spectroscopy observations of LHS 1140 b., Comment: 26 pages, 14 figures, 3 tables, accepted for publication in ApJL