Search

We show how to take the first step in the conformal program for constructing general matter couplings to Carroll gravity. In particular, we couple a single massless electric/magnetic scalar to conformal Carroll gravity with isotropic dilatations and show how, upon gauge-fixing, we obtain a (non-conformal version of) electric/magnetic Carroll gravity. We determine the full Carroll transformation rules paying special attention to the way the so-called intrinsic torsion tensors occur in these transformation rules. A noteworthy feature in the magnetic case is that the Lagrange multiplier present in the Lagrangian gets absorbed, after coupling to conformal Carroll gravity and gauge-fixing, into one of the independent spin-connections of magnetic Carroll gravity. Our results form a convenient starting point for constructing general matter couplings to Carroll gravity. Surprisingly, we find that the same relation between dynamical matter and gravity, which forms the basis of the conformal program, does not work in the usual way in the Galilei case., Comment: 28 pages

Since the derivation of a well-defined $D\to4$ limit for 4D Einstein-Gauss-Bonnet (4DEGB) gravity coupled to a scalar field, there has been considerable interest in testing it as an alternative to Einstein's general theory of relativity. Past work has shown that this theory hosts interesting compact star solutions which are smaller in radius than a Schwarzschild black hole of the same mass in general relativity (GR), though the stability of such objects has been subject to question. In this paper we solve the equations for radial perturbations of neutron stars in the 4DEGB theory with SLy/BSk class EOSs, along with the MS2 EOS, and show that the coincidence of stability and maximum mass points in GR is still present in this modified theory, with the interesting additional feature of solutions re-approaching stability near the black hole solution on the mass-radius diagram. Besides this, as expected from past work, we find that larger values of the 4DEGB coupling $\alpha$ tend to increase the mass of neutron stars of the same radius (due to a larger $\alpha$ weakening gravity) and move the maximum mass points of the solution branches closer to the black hole horizon.

Because of the rapid development and increasing public availability of Generative Artificial Intelligence (GenAI) models and tools, educational institutions and educators must immediately reckon with the impact of students using GenAI. There is limited prior research on computing students' use and perceptions of GenAI. In anticipation of future advances and evolutions of GenAI, we capture a snapshot of student attitudes towards and uses of yet emerging GenAI, in a period of time before university policies had reacted to these technologies. We surveyed all computer science majors in a small engineering-focused R1 university in order to: (1) capture a baseline assessment of how GenAI has been immediately adopted by aspiring computer scientists; (2) describe computing students' GenAI-related needs and concerns for their education and careers; and (3) discuss GenAI influences on CS pedagogy, curriculum, culture, and policy. We present an exploratory qualitative analysis of this data and discuss the impact of our findings on the emerging conversation around GenAI and education., Comment: 7 pages

Large Language Models (LLMs) store and retrieve vast amounts of factual knowledge acquired during pre-training. Prior research has localized and identified mechanisms behind knowledge recall; however, it has primarily focused on English monolingual models. The question of how these processes generalize to other languages and multilingual LLMs remains unexplored. In this paper, we address this gap by conducting a comprehensive analysis of two highly multilingual LLMs. We assess the extent to which previously identified components and mechanisms of factual recall in English apply to a multilingual context. Then, we examine when language plays a role in the recall process, uncovering evidence of language-independent and language-dependent mechanisms.

Knowledge claims are abundant in the literature on large language models (LLMs); but can we say that GPT-4 truly "knows" the Earth is round? To address this question, we review standard definitions of knowledge in epistemology and we formalize interpretations applicable to LLMs. In doing so, we identify inconsistencies and gaps in how current NLP research conceptualizes knowledge with respect to epistemological frameworks. Additionally, we conduct a survey of 100 professional philosophers and computer scientists to compare their preferences in knowledge definitions and their views on whether LLMs can really be said to know. Finally, we suggest evaluation protocols for testing knowledge in accordance to the most relevant definitions., Comment: EMNLP 2024

This paper investigates sound and music interactions arising from the use of electromyography (EMG) to instrumentalise signals from muscle exertion of the human body. We situate EMG within a family of embodied interaction modalities, where it occupies a middle ground, considered as a ''signal from the inside'' compared with external observations of the body (e.g., motion capture), but also seen as more volitional than neurological states recorded by brain electroencephalogram (EEG). To understand the messiness of gestural interaction afforded by EMG, we revisit the phenomenological turn in HCI, reading Paul Dourish's work on the transparency of ''ready-to-hand'' technologies against the grain of recent posthumanist theories, which offer a performative interpretation of musical entanglements between bodies, signals, and representations. We take music performance as a use case, reporting on the opportunities and constraints posed by EMG in workshop-based studies of vocal, instrumental, and electronic practices. We observe that across our diverse range of musical subjects, they consistently challenged notions of EMG as a transparent tool that directly registered the state of the body, reporting instead that it took on ''present-at-hand'' qualities, defamiliarising the performer's own sense of themselves and reconfiguring their embodied practice.

We present an update on the EAVI physiological interface, a wireless, microcontroller based hardware design for the acquisition of bioelectrical signals. The system has been updated to process electroencephalogram brain signals in addition to muscle electromyogram. The hardware/firmware system interfaces with host software carrying out feature extraction and signal processing. Recent advances in electronics have made physiological computing applications practical and feasible. However, there is a gap between high end biomedical equipment and consumer DIY solutions. The hardware design we present here bridges this gap, and combines a specialized biosignal acquisition chip mated with a general-purpose microcontroller. It is based on the Texas Instruments ADS129x family a single chip integrated solution for high quality biosignal amplification and digitization. It serves as analogue front end via programmable gain amplifiers to a 24bit delta-sigma analog-digital converter. The microcontroller is the STMicroelectronics STM32F427, a Cortex-M4 family microcontroller with floating point unit . In addition to EMG acquisition, the board includes a Kionix KX122 three-axis accelerometer. The TI and Kionix sensing chipts communicate with the ST microcontroller over an I2C digital serial bus. The board communicates with the host computer or rest of the music system wirelessly over Bluetooth LE 4.2 using an ST SPBTLE-1S transceiver. The board can also communicate over USB where it registers with the host as a class compliant audio and MIDI device. Audio and physiological signals are treated in the same signal processing chain using the OWL framework. The demo will show multichannel EMG, and single channel EEG. We call this hybridization ''ExG''. We will present documentation of the EAVI board used in the lab and on stage, in user studies with neuro-diverse musicians and trained instrumentalists, as well as in performance with the experimental all-female band, Chicks on Speed., Comment: NIME 2023, NIME - New Interfaces for Musical Expression, May 2023, Mexico CIty, Mexico

We report VI CCD photometry of the globular cluster cluster NGC 1851. We aim to study the membership of the variable stars detected in the field of the cluster as listed in the Catalogue of Variable stars in Globular Clusters (CVSGC; Clement et al. 2001) and reported by the Gaia mission. We cross match the two sets of variables to produce light curves that lead to the estimation of physical parameters. The resulting colour-magnitude diagram (CMD), free of likely field stars, enables to confirm the position of the variables, their type and evolutionary stage. We provide new estimations of the period using data acquired on a long timebase. The Fourier decomposition of cluster member RR Lyrae light curves lead to a mean metalicity and distance of \([Fe/H]_{ZW} = -1.35 \pm 0.22\) dex and \(11.9 \pm 0.6\) kpc. The variability and membership of stars reported by Gaia-DR3 as variables in the field of the cluster is discussed., Comment: 16 pages,10 figures, 5 table

The challenge of satellite stabilization, particularly those with uncertain flexible dynamics, has become a pressing concern in control and robotics. These uncertainties, especially the dynamics of a third-party client satellite, significantly complicate the stabilization task. This paper introduces a novel adaptive detumbling method to handle non-rigid satellites with unknown motion dynamics (translation and rotation). The distinctive feature of our approach is that we model the non-rigid tumbling satellite as a two-link serial chain with unknown stiffness and damping in contrast to previous detumbling research works which consider the satellite a rigid body. We develop a novel adaptive robotics approach to detumble the satellite by using two space tugs as servicer despite the uncertain dynamics in the post-capture case. Notably, the stiffness properties and other physical parameters, including the mass and inertia of the two links, remain unknown to the servicer. Our proposed method addresses the challenges in detumbling tasks and paves the way for advanced manipulation of non-rigid satellites with uncertain dynamics., Comment: This paper has been accepted by the 63rd IEEE Conference on Decision and Control(CDC2024) as a regular paper

Continuous-variable projective measurements can not select individual measurement results as in the discrete case; instead, the possible outcomes are bounded by the selectivity interval of the measurement; then, it is say that continuous-variable measurement devices are insufficiently selective. By utilizing this concept we show that the probability and fidelity of teleportation in a two-mode cluster state can be handled by the localization of the selectivity interval of the measurement apparatus. Besides, we provide a mathematical expression describing the probability distribution of the measurement outcomes in the two-mode cluster, which is a fundamental solution of the heat equation. In addition, we show that the fidelity of teleportation in the two mode cluster is given by the quotient between the squared solution of a non-homogeneous heat equation and the solution of the conventional heat equation. Furthermore, we extend our approach to a configuration involving successive clusters with intermediate corrections between each teleportation step. To exemplify our proposal, we consider the specific case of a squeezed-coherent state as the quantum state under teleportation., Comment: 26 pages, 8 figures

Transportation missions in aerospace are limited to the capability of each aerospace robot and the properties of the target transported object, such as mass, inertia, and grasping locations. We present a novel decentralized adaptive controller design for multiple robots that can be implemented in different kinds of aerospace robots. Our controller adapts to unknown objects in different gravity environments. We validate our method in an aerial scenario using multiple fully actuated hexarotors with grasping capabilities, and a space scenario using a group of space tugs. In both scenarios, the robots transport a payload cooperatively through desired three-dimensional trajectories. We show that our method can adapt to unexpected changes that include the loss of robots during the transportation mission., Comment: This paper has been accepted by DARS2024 Conference. The permission for the preprint version on Arxiv has been approved through the DARS2024 Committee and Springer Press

We made a survey of the variable stars in a $13.2 \times 13.2$ arcmin$^2$ centered on the field of the Galactic bulge cluster NGC 6558. A total of 78 variables was found in the field of the cluster. Many of these variables are included in the Catalogue of Variable Stars in Galactic Globular Clusters (Clement et al. 2001), OGLE or Gaia-DR3 data releases. A membership analysis based on the proper motions of Gaia-DR3 revealed that many of these variables do not belong to the cluster. We employed the data from the aforementioned surveys and our own data in the VI photometric system to estimate the periods, which along with the light curves morphology and position in a deferentially dereddened colour-magnitude diagram(CMD), help classifying the variable types. Two new member variables were found; an eclipsing binary (V18) and a semi-regular SR/L (V19). In the end we conclude that only 9 variables are likely cluster members. Member variables were used to discuss the mean metallicity and distance of the parental cluster and find the average values.

Employing \emph{VI} images of NGC 2419 acquired over 17 years, light curves for most of the known variables in the field of the cluster are produced. A cluster membership analysis for about 3100 stars in the cluster field with proper motions from $Gaia$-DR3, revealed the presence of member stars as far as 140 pc from the cluster center and enabled the construction of a cleaner CMD free of field stars. It was found that RRab and RRc stars share the inter-order region in the instability strip, which is unusual for OoII clusters. Theoretical considerations confirm that Pop II cepheids are descendants of extreme ZAHB blue tail stars with very thin envelopes of about 10\% of the total mass. Member RR Lyrae stars were employed to calculate independent estimates of the mean cluster metallicity and distance; we found [Fe/H]$_{\rm UV}= -1.90 \pm 0.27$ and $D=86.3 \pm 5.0$ kpc from the RRab and [Fe/H]$_{\rm UV}= -1.88 \pm 0.30$ and $D=83.1 \pm 8.1$ kpc from the RRc light curves., Comment: 19 pages, 11 figures, 3 tables

The purpose of instruction tuning is enabling zero-shot performance, but instruction tuning has also been shown to improve chain-of-thought reasoning and value alignment (Si et al., 2023). Here we consider the impact on $\textit{consistency}$, i.e., the sensitivity of language models to small perturbations in the input. We compare 10 instruction-tuned LLaMA models to the original LLaMA-7b model and show that almost across-the-board they become more consistent, both in terms of their representations and their predictions in zero-shot and downstream tasks. We explain these improvements through mechanistic analyses of factual recall., Comment: We need to run extra experiments to ensure some of the claims in the paper are fully correct

The increasing demand for the deployment of LLMs in information-seeking scenarios has spurred efforts in creating verifiable systems, which generate responses to queries along with supporting evidence. In this paper, we explore the attribution capabilities of plan-based models which have been recently shown to improve the faithfulness, grounding, and controllability of generated text. We conceptualize plans as a sequence of questions which serve as blueprints of the generated content and its organization. We propose two attribution models that utilize different variants of blueprints, an abstractive model where questions are generated from scratch, and an extractive model where questions are copied from the input. Experiments on long-form question-answering show that planning consistently improves attribution quality. Moreover, the citations generated by blueprint models are more accurate compared to those obtained from LLM-based pipelines lacking a planning component., Comment: Accepted at ACL 2024

Facts are subject to contingencies and can be true or false in different circumstances. One such contingency is time, wherein some facts mutate over a given period, e.g., the president of a country or the winner of a championship. Trustworthy language models ideally identify mutable facts as such and process them accordingly. We create MuLan, a benchmark for evaluating the ability of English language models to anticipate time-contingency, covering both 1:1 and 1:N relations. We hypothesize that mutable facts are encoded differently than immutable ones, hence being easier to update. In a detailed evaluation of six popular large language models, we consistently find differences in the LLMs' confidence, representations, and update behavior, depending on the mutability of a fact. Our findings should inform future work on the injection of and induction of time-contingent knowledge to/from LLMs.

This article studies the physical and electrical behavior of Gd2-xScxO3 layers grown by high pressure sputtering from metallic Gd and Sc targets. The aim is to obtain a high permittivity dielectric for microelectronic applications. The films were obtained by the deposition of a metallic nanolaminate of Gd and Sc alternating layers, which is afterwards in-situ oxidized by plasma. The oxide films obtained were close to stoichiometry, amorphous and with minimal interfacial regrowth. By fabricating metal-insulator-semiconductor capacitors we found that a moderate temperature annealing is needed to enhance permittivity, which reaches a high value of 32 while keeping moderate leakage. Finally, the feasibility of interface scavenging in this material with Ti gate electrodes is also demonstrated., Comment: 36 pages, 13 figures

Gd-rich gadolinium scandate (Gd2-xScxO3) was deposited by high-pressure sputtering on (100) silicon by alternating the deposition of lower than 0.5 nm thick films of its binary components: Sc2O3 and Gd2O3. The formation of the ternary oxide was observed after the thermal treatments, with a high increase in the effective permittivity of the dielectric (up to 21). The silicon diffuses into the Gd2-xScxO3 films, which show an amorphous character. After the annealing no interfacial silicon oxide is present. CHF-VG curves indicated low hysteresis (55 mV) and a density of interfacial defects of 6e11 eV-1 cm-2., Comment: 25 pages, 9 figures, 2 tables

In this paper we present the Hietarinta Chern-Simons supergravity theory in three space-time dimensions which extends the simplest Poincar\'e supergravity theory. After approaching the construction of the action using the Chern-Simons formalism, the analysis of the corresponding asymptotic symmetry algebra is considered. For this purpose, we first propose a consistent set of asymptotic boundary conditions for the aforementioned supergravity theory whose underlying symmetry corresponds to the supersymmetric extension of the Hietarinta algebra. We then show that the corresponding charge algebra contains the super-$\mathfrak{bms}_{3}$ algebra as subalgebra, and has three independent central charges. We also show that the obtained asymptotic symmetry algebra can alternatively be recovered as a vanishing cosmological constant limit of three copies of the Virasoro algebra, one of which is augmented by supersymmetry., Comment: 20 pages

This letter introduces an innovative method to enhance the quality of audio time stretching by precisely decomposing a sound into sines, transients, and noise and by improving the processing of the latter component. While there are established methods for time-stretching sines and transients with high quality, the manipulation of noise or residual components has lacked robust solutions in prior research. The proposed method combines sound decomposition with previous techniques for audio spectral resynthesis. The time-stretched noise component is achieved by morphing its time-interpolated spectral magnitude with a white-noise excitation signal. This method stands out for its simplicity, efficiency, and audio quality. The results of a subjective experiment affirm the superiority of this approach over current state-of-the-art methods across all evaluated stretch factors. The proposed technique notably excels in extreme stretching scenarios, signifying a substantial elevation in performance. The proposed method holds promise for a wide range of applications in slow-motion media content, such as music or sports video production., Comment: submitted to IEEE Signal Processing Letters

In the past 15 years, two international observatories have been intensively studying entrepreneurship using empirical studies with different methodologies: GEM and PSED. Both projects have generated a considerable volume of scientific production, and their intellectual structures are worth analyzing. The current work is an exploratory study of the knowledge base of the articles generated by each of these two observatories and published in prestigious journals. The value added of this work lies in its novel characterization of the intellectual structure of entrepreneurship according to the academic production of these two initiatives. The results may be of interest to the managers and members of these observatories, as well as to academics, researchers, sponsors and policymakers interested in entrepreneurship.

We examine the minimal constraints imposed by the Weak Gravity Conjecture (WGC) on the particle spectrum of a quantum gravity theory. Towers of super-extremal states have previously been argued to be required for consistency of the WGC under circle reduction. At the same time, there exist classes of theories where no tower of super-extremal particle states below the black hole threshold has been established with current techniques. We resolve this tension by arguing for the existence of a minimal radius for circle reductions of generic quantum gravity theories. Below this threshold, the notion of a circle compactification breaks down, bypassing the need for a tower of super-extremal states to satisfy the WGC after circle reduction. Based on this we propose that if a theory satisfies the WGC at the particle level below the black hole threshold, these states are sufficient for consistency under dimensional reduction, even in absence of a tower of super-extremal particles. Apart from general arguments, we provide independent evidence for this main result in F-, M- and string theory compactifications. According to the Emergent String Conjecture the only exception to the appearance of a minimal radius arises in asymptotically weak-coupling limits for heterotic strings, which aligns with the appearance of a weakly coupled super-extremal tower of particle states. This observation motivates a Minimal Weak Gravity Conjecture which states that towers of super-extremal particles occur if and only if they are required by consistency of the WGC under dimensional reduction., Comment: v2: 62 (55+appendix) pages, JHEP version

We compute the bulk, $\zeta$, and shear, $\eta$, viscosity over entropy density, $s$, for the QCD matter formed in small collision systems at LHC. We consider a scenario of the String Percolation Model by proposing a global form of the color reduction factor that describes both the thermodynamic limit and its maximum deviation due to small-bounded effects. Our method involves estimations at vanishing baryon-chemical potential, assuming local equilibrium for string clusters in the initial state. To compute $\eta/s$, we employed a kinetic approach that accounts QCD states as an ideal gas of partons, while $\zeta/s$ is computed by using two different approaches: a simple kinetic formula and the causal dissipative relativistic fluid dynamics formulation. Our results align with Lattice QCD computations and Bayesian methods and are consistent with holographic conjecture bounds. Furthermore, our findings support the notion of a strongly interacting medium, similar to that observed in nuclear collisions, albeit with a phase transition occurring outside the thermodynamic limit., Comment: 12 pages, 8 figures