Datasets:

librarian-bots

/

arxiv-metadata-snapshot

like 3

2312.05475

Kiyoshi Kanazawa

Kiyoshi Kanazawa and Didier Sornette

A standard form of master equations for general non-Markovian jump processes: the Laplace-space embedding framework and asymptotic solution

29 pages, 8 figures

Phys. Rev. Res. 6, 023270 (2024)

10.1103/PhysRevResearch.6.023270

cond-mat.stat-mech physics.data-an q-fin.TR

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We present a standard form of master equations (ME) for general one-dimensional non-Markovian (history-dependent) jump processes, complemented by an asymptotic solution derived from an expanded system-size approach. The ME is obtained by developing a general Markovian embedding using a suitable set of auxiliary field variables. This Markovian embedding uses a Laplace-convolution operation applied to the velocity trajectory. We introduce an asymptotic method tailored for this ME standard, generalising the system-size expansion for these jump processes. Under specific stability conditions tied to a single noise source, upon coarse-graining, the Generalized Langevin Equation (GLE) emerges as a universal approximate model for point processes in the weak-coupling limit. This methodology offers a unified analytical toolset for general non-Markovian processes, reinforcing the universal applicability of the GLE founded in microdynamics and the principles of statistical physics.

2312.05573

Ayoub Belhadji

Ayoub Belhadji and R\'emi Gribonval

Revisiting RIP guarantees for sketching operators on mixture models

stat.ML cs.LG

http://creativecommons.org/licenses/by/4.0/

In the context of sketching for compressive mixture modeling, we revisit existing proofs of the Restricted Isometry Property of sketching operators with respect to certain mixtures models. After examining the shortcomings of existing guarantees, we propose an alternative analysis that circumvents the need to assume importance sampling when drawing random Fourier features to build random sketching operators. Our analysis is based on new deterministic bounds on the restricted isometry constant that depend solely on the set of frequencies used to define the sketching operator; then we leverage these bounds to establish concentration inequalities for random sketching operators that lead to the desired RIP guarantees. Our analysis also opens the door to theoretical guarantees for structured sketching with frequencies associated to fast random linear operators.

2312.05812

Yuriy Pogorelov

Y.G. Pogorelov, V.M. Loktev

Topological conditions for impurity effects in graphene nanosystems

14 pages, 17 figures

cond-mat.mes-hall

http://creativecommons.org/licenses/by/4.0/

We consider electronic spectra of graphene nanotubes and their perturbation by impurity atoms absorbed at different positions on nanotube surfaces, within the framework of Anderson hybrid model. A special attention is given to the cases when Dirac-like 1D modes appear in the nanotube spectrum and their hybridization with localized impurity states produces, at growing impurity concentration $c$, onset of a mobility gap near the impurity level and even opening, at yet higher $c$, of some narrow delocalized range within this mobility gap. Such behaviors are compared with the similar effects in the previously studied 2D graphene and armchair type graphene nanoribbons. Some possible practical applications are discussed.

2312.06112

Jaeseok Byun

Jaeseok Byun, Dohoon Kim, and Taesup Moon

MAFA: Managing False Negatives for Vision-Language Pre-training

CVPR 2024 camera ready version

cs.CV cs.AI

http://creativecommons.org/licenses/by-nc-sa/4.0/

We consider a critical issue of false negatives in Vision-Language Pre-training (VLP), a challenge that arises from the inherent many-to-many correspondence of image-text pairs in large-scale web-crawled datasets. The presence of false negatives can impede achieving optimal performance and even lead to a significant performance drop. To address this challenge, we propose MAFA (MAnaging FAlse negatives), which consists of two pivotal components building upon the recently developed GRouped mIni-baTch sampling (GRIT) strategy: 1) an efficient connection mining process that identifies and converts false negatives into positives, and 2) label smoothing for the image-text contrastive (ITC) loss. Our comprehensive experiments verify the effectiveness of MAFA across multiple downstream tasks, emphasizing the crucial role of addressing false negatives in VLP, potentially even surpassing the importance of addressing false positives. In addition, the compatibility of MAFA with the recent BLIP-family model is also demonstrated. Code is available at https://github.com/jaeseokbyun/MAFA.

2312.07273

Dinh Tuan Truong

Tuan Truong, Farnaz Khun Jush, Matthias Lenga

Benchmarking Pretrained Vision Embeddings for Near- and Duplicate Detection in Medical Images

Accepted to International Symposium on Biomedical Imaging 2024, Athens, Greece

cs.CV

http://creativecommons.org/licenses/by/4.0/

Near- and duplicate image detection is a critical concern in the field of medical imaging. Medical datasets often contain similar or duplicate images from various sources, which can lead to significant performance issues and evaluation biases, especially in machine learning tasks due to data leakage between training and testing subsets. In this paper, we present an approach for identifying near- and duplicate 3D medical images leveraging publicly available 2D computer vision embeddings. We assessed our approach by comparing embeddings extracted from two state-of-the-art self-supervised pretrained models and two different vector index structures for similarity retrieval. We generate an experimental benchmark based on the publicly available Medical Segmentation Decathlon dataset. The proposed method yields promising results for near- and duplicate image detection achieving a mean sensitivity and specificity of 0.9645 and 0.8559, respectively.

2312.08147

Mario Fuest

Mario Fuest and Shahin Heydari

A cross-diffusion system modelling rivaling gangs: global existence of bounded solutions and FCT stabilization for numerical simulation

34 pages, 20 figures

10.1142/S0218202524500349

math.AP cs.NA math.NA

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

For the gang territoriality model \begin{align*} \begin{cases} u_t = D_u \Delta u + \chi_u \nabla \cdot (u \nabla w), \\ v_t = D_v \Delta v + \chi_v \nabla \cdot (v \nabla z), \\ w_t = -w + \frac{v}{1+v}, \\ z_t = -z + \frac{u}{1+u}, \end{cases} \end{align*} where $u$ and $v$ denote the densities of two rivaling gangs which spray graffiti (with densities $z$ and $w$, respectively) and partially move away from the other gang's graffiti, we construct global, bounded classical solutions. By making use of quantitative global estimates, we prove that these solutions converge to homogeneous steady states if $\|u_0\|_{L^\infty(\Omega)}$ and $\|v_0\|_{L^\infty(\Omega)}$ are sufficiently small. Moreover, we perform numerical experiments which show that for different choices of parameters, the system may become diffusion- or convection-dominated, where in the former case the solutions converge toward constant steady states while in the later case nontrivial asymptotic behavior such as segregation is observed. In order to perform these experiments, we apply a nonlinear finite element flux-corrected transport method (FEM-FCT) which is positivity-preserving. Then, we treat the nonlinearities in both the system and the proposed nonlinear scheme simultaneously using fixed-point iteration.

2312.09292

Jack Y. Araz

Jack Y. Araz, Michael Spannowsky and Matthew Wingate

Exploring thermal equilibria of the Fermi-Hubbard model with variational quantum algorithms

8 pages, 5 figures. Accepted version for publication in PRA

IPPP/23/72, JLAB-THY-23-3968

quant-ph cond-mat.mtrl-sci hep-lat

http://creativecommons.org/licenses/by/4.0/

This study investigates the thermal properties of the repulsive Fermi-Hubbard model with chemical potential using variational quantum algorithms, crucial in comprehending particle behaviour within lattices at high temperatures in condensed matter systems. Conventional computational methods encounter challenges, especially in managing chemical potential, prompting exploration into Hamiltonian approaches. Despite the promise of quantum algorithms, their efficacy is hampered by coherence limitations when simulating extended imaginary time evolution sequences. To overcome such constraints, this research focuses on optimising variational quantum algorithms to probe the thermal properties of the Fermi-Hubbard model. Physics-inspired circuit designs are tailored to alleviate coherence constraints, facilitating a more comprehensive exploration of materials at elevated temperatures. Our study demonstrates the potential of variational algorithms in simulating the thermal properties of the Fermi-Hubbard model while acknowledging limitations stemming from error sources in quantum devices and encountering barren plateaus.

2312.09479

Takashi Ui

Masaki Miyashita, Takashi Ui

LQG Information Design

econ.TH cs.SY eess.SY

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

This paper addresses information design with linear best responses of agents, quadratic objective functions of an information designer, and a payoff state distributed according to a Gaussian distribution. We formulate the problem as semidefinite programming (SDP) and use the duality principle to characterize an optimal information structure. There exists a Gaussian information structure that is optimal among all information structures. A necessary and sufficient condition for optimality is that the realizations of the induced action profile and a state satisfy linear constraints derived from the primal and dual SDP. As a result, an observed action profile typically reveals the true state even if individual agents have only partial knowledge. In symmetric network games, an optimal information structure inherits this symmetry, which facilitates the computation of an optimal information structure.

2312.10576

Lei Zhang

Lei Zhang, Emanuel Gull

Minimal Pole Representation and Controlled Analytic Continuation of Matsubara Response Functions

cond-mat.str-el physics.comp-ph

http://creativecommons.org/licenses/by/4.0/

Analytical continuation is a central step in the simulation of finite-temperature field theories in which numerically obtained Matsubara data is continued to the real frequency axis for physical interpretation. Numerical analytic continuation is considered to be an ill-posed problem where uncertainties on the Matsubara axis are aplified exponentially. Here, we present a systematic and controlled procedure that approximates any Matsubara function by a minimal pole representation to within a predefined precision. We then show systematic convergence to the exact spectral function on the real axis as a function of our control parameter for a range of physically relevant setups. Our methodology is robust to noise and paves the way towards reliable analytic continuation in many-body theory and, by providing access to the analytic structure of the functions, direct theoretical interpretation of physical properties.

2312.10845

Chen Wan

Rapha\"el Beuzart-Plessis, Chen Wan

A local twisted trace formula for Whittaker induction of coregular symmetric pairs: the geometric side

137 pages, comments are welcome

math.RT math.NT

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

In this paper, we prove the geometric expansion of a local twisted trace formula for the Whittaker induction of any symmetric pairs that are coregular. This generalizes the local (twisted) trace formula for reductive groups proved by Arthur \cite{A91} and Waldspurger \cite{WalFTLtordue}. We also prove a formula for the regular germs of quasi-characters associated to strongly cuspidal functions in terms of certain weighted orbital integrals. As a consequence of our trace formula and the formula for regular germs of quasi-characters, we prove a simple local trace formula of those models for strongly cuspidal test functions which implies a multiplicity formula for these models. We also present various applications of our trace formula and multiplicity formula, including a necessary condition for a discrete L-packet to contain a representation with a unitary Shalika model (resp. a Galois model for classical groups) in terms of the associated Langlands parameter, and we also compute the summation of the corresponding multiplicities for certain discrete L-packets.

2312.11326

Marcelo Locatelli

Marcelo Sartori Locatelli, Pedro Calais, Matheus Prado Miranda, Jo\~ao Pedro Junho, Tomas Lacerda Muniz, Wagner Meira Jr., Virgilio Almeida

Topic Shifts as a Proxy for Assessing Politicization in Social Media

12 pages, 6 figures, accepted for the 18th AAAI International Conference on Web and Social Media (ICWSM-2024)

Topic Shifts as a Proxy for Assessing Politicization in Social Media. In: Proceedings of the International AAAI Conference on Web and Social Media. 2024. p. 972-984

10.1609/icwsm.v18i1.31366

cs.SI

http://creativecommons.org/licenses/by/4.0/

Politicization is a social phenomenon studied by political science characterized by the extent to which ideas and facts are given a political tone. A range of topics, such as climate change, religion and vaccines has been subject to increasing politicization in the media and social media platforms. In this work, we propose a computational method for assessing politicization in online conversations based on topic shifts, i.e., the degree to which people switch topics in online conversations. The intuition is that topic shifts from a non-political topic to politics are a direct measure of politicization -- making something political, and that the more people switch conversations to politics, the more they perceive politics as playing a vital role in their daily lives. A fundamental challenge that must be addressed when one studies politicization in social media is that, a priori, any topic may be politicized. Hence, any keyword-based method or even machine learning approaches that rely on topic labels to classify topics are expensive to run and potentially ineffective. Instead, we learn from a seed of political keywords and use Positive-Unlabeled (PU) Learning to detect political comments in reaction to non-political news articles posted on Twitter, YouTube, and TikTok during the 2022 Brazilian presidential elections. Our findings indicate that all platforms show evidence of politicization as discussion around topics adjacent to politics such as economy, crime and drugs tend to shift to politics. Even the least politicized topics had the rate in which their topics shift to politics increased in the lead up to the elections and after other political events in Brazil -- an evidence of politicization.

2312.11660

Ant\'onio Antunes

Ant\'onio Antunes, Sebastian Harris, Apratim Kaviraj, Volker Schomerus

Lining up a Positive Semi-Definite Six-Point Bootstrap

45 pages + appendices, 13 figures; v2: added references, minor clarifications and details on numerical implementation, matches JHEP version

JHEP 2024, 58 (2024)

10.1007/JHEP06(2024)058

DESY-23-220

hep-th

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

In this work we initiate a positive semi-definite numerical bootstrap program for multi-point correlators. Considering six-point functions of operators on a line we reformulate the crossing symmetry equation for a pair of comb-channel expansions as a semi-definite programming problem. We provide two alternative formulations of this problem. At least one of them turns out to be amenable to numerical implementation. Through a combination of analytical and numerical techniques we obtain rigorous bounds on CFT data in the triple-twist channel for several examples.

2312.12794

Yifan Wang

Sahil Singla, Yifan Wang

Bandit Sequential Posted Pricing via Half-Concavity

cs.LG cs.DS cs.GT

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Sequential posted pricing auctions are popular because of their simplicity in practice and their tractability in theory. A usual assumption in their study is that the Bayesian prior distributions of the buyers are known to the seller, while in reality these priors can only be accessed from historical data. To overcome this assumption, we study sequential posted pricing in the bandit learning model, where the seller interacts with $n$ buyers over $T$ rounds: In each round the seller posts $n$ prices for the $n$ buyers and the first buyer with a valuation higher than the price takes the item. The only feedback that the seller receives in each round is the revenue. Our main results obtain nearly-optimal regret bounds for single-item sequential posted pricing in the bandit learning model. In particular, we achieve an $\tilde{O}(\mathsf{poly}(n)\sqrt{T})$ regret for buyers with (Myerson's) regular distributions and an $\tilde{O}(\mathsf{poly}(n)T^{{2}/{3}})$ regret for buyers with general distributions, both of which are tight in the number of rounds $T$. Our result for regular distributions was previously not known even for the single-buyer setting and relies on a new half-concavity property of the revenue function in the value space. For $n$ sequential buyers, our technique is to run a generalized single-buyer algorithm for all the buyers and to carefully bound the regret from the sub-optimal pricing of the suffix buyers.

2312.13051

Alexander Tyner

Alexander C. Tyner

BerryEasy: A GPU enabled python package for diagnosis of nth-order and spin-resolved topology in the presence of fields and effects

10 pages, 7 figures, published version, code updated

J. Phys.: Condens. Matter 36 325902, 2024

10.1088/1361-648X/ad475f

cond-mat.mtrl-sci cond-mat.dis-nn cond-mat.mes-hall cond-mat.str-el

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Multiple software packages currently exist for the computation of bulk topological invariants in both idealized tight-binding models and realistic Wannier tight-binding models derived from density functional theory. Currently, only one package, PythTB(https://www.physics.rutgers.edu/pythtb/) is capable of computing nested Wilson loops and spin-resolved Wilson loops. These state-of-the-art techniques are vital for accurate analysis of band topology. In this paper we introduce BerryEasy, a python package which is built to work alongside the PyBinding(https://docs.pybinding.site/en/stable/index.html) software package. By working in tandem with the Pybinding package and harnessing the speed of graphical processing units, topological analysis of supercells in the presence of disorder and impurities is made possible. The BerryEasy package simultaneously accommodates use of realistic tight-binding models developed using Wannier90.

2312.13267

Ricardo Monteiro

George Doran, Ricardo Monteiro, Sam Wikeley

On the anomaly interpretation of amplitudes in self-dual Yang-Mills and gravity

41 pages, 2 figures. v2: published version with substantial improvements, including new section 4

QMUL-PH-23-26, UUITP--38/23

hep-th

http://creativecommons.org/licenses/by/4.0/

We investigate the integrability anomalies arising in the self-dual sectors of gravity and Yang-Mills theory, focusing on their connection to both the chiral anomaly and the trace anomaly. The anomalies in the self-dual sectors generate the one-loop all-plus amplitudes of gravitons and gluons, and have recently been studied via twistor constructions. On the one hand, we show how they can be interpreted as an anomaly of the chiral U(1) electric-magnetic-type duality in the self-dual sectors. We also note the similarity, for the usual fermionic chiral anomaly, between the 4D setting of self-dual Yang-Mills and the 2D setting of the Schwinger model. On the other hand, the anomalies in the self-dual theories also resemble the trace anomaly, sharing the same type of non-local effective action. We highlight the role of a Weyl-covariant fourth-order differential operator familiar from the trace anomaly literature, which (i) explains the conformal properties of the one-loop amplitudes, and (ii) indicates how this story may be extended to non-trivial spacetime backgrounds, e.g. with a cosmological constant. Moving beyond the self-dual sectors, and focusing on the gravity case, we comment on an intriguing connection to the two-loop ultraviolet divergence of pure gravity, whereby cancelling the anomaly at one-loop eliminates the two-loop divergence for the simplest helicity amplitudes.

2312.13274

Michael Brown

Michael D. Brown, Adam Meily, Brian Fairservice, Akshay Sood, Jonathan Dorn, Eric Kilmer, Ronald Eytchison

A Broad Comparative Evaluation of Software Debloating Tools

17 pages, 8 tables

cs.SE cs.CR cs.PL

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Software debloating tools seek to improve program security and performance by removing unnecessary code, called bloat. While many techniques have been proposed, several barriers to their adoption have emerged. Namely, debloating tools are highly specialized, making it difficult for adopters to find the right type of tool for their needs. This is further hindered by a lack of established metrics and comparative evaluations between tools. To close this information gap, we surveyed 10 years of debloating literature and several tools currently under commercial development to taxonomize knowledge about the debloating ecosystem. We then conducted a broad comparative evaluation of 10 debloating tools to determine their relative strengths and weaknesses. Our evaluation, conducted on a diverse set of 20 benchmark programs, measures tools across 12 performance, security, and correctness metrics. Our evaluation surfaces several concerning findings that contradict the prevailing narrative in the debloating literature. First, debloating tools lack the maturity required to be used on real-world software, evidenced by a slim 22% overall success rate for creating passable debloated versions of medium- and high-complexity benchmarks. Second, debloating tools struggle to produce sound and robust programs. Using our novel differential fuzzing tool, DIFFER, we discovered that only 13% of our debloating attempts produced a sound and robust debloated program. Finally, our results indicate that debloating tools typically do not improve the performance or security posture of debloated programs by a significant degree according to our evaluation metrics. We believe that our contributions in this paper will help potential adopters better understand the landscape of tools and will motivate future research and development of more capable debloating tools.

2312.13283

Marcia Tenser

Luigi Castiglioni, Silvia Penati, Marcia Tenser, Diego Trancanelli

Interpolating Bremsstrahlung function in ABJM

27 pages + appendices, 5 figures

10.1103/PhysRevD.109.126010

hep-th

http://creativecommons.org/licenses/by/4.0/

In ABJM theory, enriched RG flows between circular 1/6 BPS bosonic and 1/2 BPS fermionic Wilson loops have been introduced in arXiv:2211.16501. These flows are triggered by deformations corresponding to parametric 1/6 BPS fermionic loops. In this paper we revisit the study of these operators, but instead of circular contours we consider an interpolating cusped line and a latitude and study their RG flow in perturbation theory. This allows for the definition of a Bremsstrahlung function away from fixed points. We generalize to this case the known cusp/latitude correspondence that relates the Bremsstrahlung function to a latitude Wilson loop. We find that away from the conformal fixed points the ordinary identity is broken by the conformal anomaly in a controlled way. From a defect perspective, the breaking of the correspondence can be traced back to the appearance of an anomalous dimension for fermionic operators localized on the defect. As a by-product, we provide a brand new result for the two-loop cusp anomalous dimension of the 1/6 BPS fermionic and the 1/6 BPS bosonic Wilson lines.

2312.14088

Alessio D'Al\`i

Alessio D'Al\`i and Emanuele Delucchi

Equivariant Hilbert and Ehrhart series under translative group actions

v2. *Major* changes, including the title; new content added, at least one mistake fixed (condition (iii) in Lemma 2.4 from v1 was strictly weaker than the others; we thank Alan Stapledon). 33 pages, 3 figures. Comments are still very welcome!

math.CO math.AC math.RT

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We study the equivariant Hilbert series of the complex Stanley-Reisner ring of a simplicial complex $\Sigma$ under the action of a finite group via simplicial automorphisms. We prove that, when $\Sigma$ is Cohen-Macaulay and the group action is translative, i.e., color-preserving for some proper coloring of $\Sigma$, the equivariant Hilbert series of $\Sigma$ admits a rational expression whose numerator is a positive integer combination of irreducible characters. We relate such a result to equivariant Ehrhart theory via an equivariant analogue of the Betke-McMullen theorem developed by Jochemko, Katth\"an and Reiner: namely, if a lattice polytope admits a lattice unimodular triangulation that is invariant under the action of a finite group, then the equivariant Ehrhart series of the polytope equals the equivariant Hilbert series of such a triangulation. As an application, we study the equivariant Ehrhart series of alcoved polytopes in the sense of Lam and Postnikov and derive explicit results in the case of order polytopes and of Lipschitz poset polytopes.

2312.15522

Sergei Adler

Sergei Adler, Hermann Boos

Fermionic Basis in Conformal Field Theory and Thermodynamic Bethe Ansatz for Excited States II

22 pages

hep-th

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We consider the XXZ spin chain in the scaling limit in the Matsubara direction. The main result of this paper is new representations for the functions $\Psi(l, \kappa)$ and $\Theta(l, m; \kappa, \alpha)$ associated with the function $\omega(\zeta, \xi; \kappa, \kappa)$ found in the expression for the correlation function of the generators of the fermionic basis for the XXZ spin chain. The final result incorporates the case of particle-hole excitations which is needed for the relation of the fermionic basis to the Virasoro basis of the CFT descendants.

2312.15538

Yinuo Du

Yinuo Du, Hanying Zhao, Yang Liu, Xinlei Yu, Yuan Shen

Exploiting Multipath Information for Integrated Localization and Sensing via PHD Filtering

This work has been submitted to the IEEE Transactions on Vehicular Technology

eess.SP

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Accurate localization and perception are pivotal for enhancing the safety and reliability of vehicles. However, current localization methods suffer from reduced accuracy when the line-of-sight (LOS) path is obstructed, or a combination of reflections and scatterings is present. In this paper, we present an integrated localization and sensing method that delivers superior performance in complex environments while being computationally efficient. Our method uniformly leverages various types of multipath components (MPCs) through the lens of random finite sets (RFSs), encompassing reflections, scatterings, and their combinations. This advancement eliminates the need for the multipath identification step and streamlines the filtering process by removing the necessity for distinct filters for different multipath types, a requirement that was critical in previous research. The simulation results demonstrate the superior performance of our method in both robustness and effectiveness, particularly in complex environments where the LOS MPC is obscured and in situations involving clutter and missed detection of MPC measurements.

2312.15950

Herve Andres

Herv\'e Andr\`es (CERMICS), Alexandre Boumezoued, Benjamin Jourdain (CERMICS, MATHRISK)

Implied volatility (also) is path-dependent

q-fin.CP

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We propose a new model for the coherent forecasting of both the implied volatility surfaces and the underlying asset returns.In the spirit of Guyon and Lekeufack (2023) who are interested in the dependence of volatility indices (e.g. the VIX) on the paths of the associated equity indices (e.g. the S&P 500), we first study how implied volatility can be predicted using the past trajectory of the underlying asset price. Our empirical study reveals that a large part of the movements of the at-the-money-forward implied volatility for up to two years maturities can be explained using the past returns and their squares. Moreover, we show that up to four years of the past evolution of the underlying price should be used for the prediction and that this feedback effect gets weaker when the maturity increases. Building on this new stylized fact, we fit to historical data a parsimonious version of the SSVI parameterization (Gatheral and Jacquier, 2014) of the implied volatility surface relying on only four parameters and show that the two parameters ruling the at-the-money-forward implied volatility as a function of the maturity exhibit a path-dependent behavior with respect to the underlying asset price. Finally, we propose a model for the joint dynamics of the implied volatility surface and the underlying asset price. The latter is modelled using a variant of the path-dependent volatility model of Guyon and Lekeufack and the former is obtained by adding a feedback effect of the underlying asset price onto the two parameters ruling the at-the-money-forward implied volatility in the parsimonious SSVI parameterization and by specifying a hidden semi-Markov diffusion model for the residuals of these two parameters and the two other parameters. Thanks to this model, we are able to simulate highly realistic paths of implied volatility surfaces that are arbitrage-free.

2312.16491

Antonio Alex-Amor

Salvador Moreno-Rodr\'iguez, Antonio Alex-Amor, Pablo Padilla, Juan F. Valenzuela-Vald\'es, Carlos Molero

Space-time Metallic Metasurfaces for Frequency Conversion and Beamforming

Physical Review Applied 21, 064018, 2024

10.1103/PhysRevApplied.21.064018

physics.app-ph

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

This paper details a class of metal-based space-time metasurfaces for application in wireless communications scenarios. Concretely, we describe space-time metasurfaces that periodically alternate their properties in time between three spatial states: "air", "conductor" and "grating". We analyze the physics of these metastructures via a computationally-efficient analytical technique based on the use of Floquet-Bloch series, integral equations and circuit models. By doing so, we reveal important features of these spatiotemporal metasurfaces: scattering parameters, field profiles, diffraction angles and nature of the space-time harmonics. The results, corroborated with a self-implemented numerical FDTD approach, show the potential application of these space-time metasurfaces as beamformers acting in reflection, in transmission or both. The amplitude and direction of the diffracted orders can be electronically controlled with the paramaters of the metasurface. Moreover, the intrinsic ability of time-modulated diffractive metasurfaces to mix and multiply frequencies is tested. We show how two different modulations can lead to the same diffraction angle but with different mixed output frequencies.

2312.16708

S\"oren Nagel

S\"oren Nagel, Jobst Heitzig and Eckehard Sch\"oll

Capital Inequality Induced Business Cycles

11 pages, 7 figures

physics.soc-ph math.DS

http://creativecommons.org/licenses/by/4.0/

In this letter we present a stochastic dynamic model which can explain economic cycles. We show that the macroscopic description yields a complex dynamical landscape consisting of multiple stable fixed points, each corresponding to a split of the population into a large low and a small high income group. The stochastic fluctuations induce switching between the resulting metastable states, and excitation oscillations just below a deterministic bifurcation. The shocks are caused by the decisions of a few agents who have a disproportionate influence over the macroscopic state of the economy due to the unequal distribution of wealth among the population. The fluctuations have a long-term effect on the growth of economic output and lead to business cycle oscillations exhibiting coherence resonance, where the correlation time is controlled by the population size which is inversely proportional to the noise intensity.

2401.00072

Jarek Paw{\l}owski

Pawe{\l} Leszczy\'nski, Kamil Kutorasi\'nski, Marcin Szewczyk, and Jaros{\l}aw Paw{\l}owski

Machine-learned models for magnetic materials

9 pages, 6 figures

cond-mat.mtrl-sci cs.LG

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We present a general framework for modeling power magnetic materials characteristics using deep neural networks. Magnetic materials represented by multidimensional characteristics (that mimic measurements) are used to train the neural autoencoder model in an unsupervised manner. The encoder is trying to predict the material parameters of a theoretical model, which is then used in a decoder part. The decoder, using the predicted parameters, reconstructs the input characteristics. The neural model is trained to capture a synthetically generated set of characteristics that can cover a broad range of material behaviors, leading to a model that can generalize on the underlying physics rather than just optimize the model parameters for a single measurement. After setting up the model, we prove its usefulness in the complex problem of modeling magnetic materials in the frequency and current (out-of-linear range) domains simultaneously, for which we use measured characteristics obtained for frequency up to $10$ MHz and H-field up to saturation.

2401.00449

Swaroop Joshi

Parthasarathy PD and Swaroop Joshi

Teaching Digital Accessibility to Industry Professionals using the Community of Practice Framework: An Experience Report

To be published in International Conference on Software Engineering (ICSE'24), Software Engineering Education and Training Track

10.1145/3639474.3640083

cs.SE cs.CY

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Despite recent initiatives aimed at improving accessibility, the field of digital accessibility remains markedly behind contemporary advancements in the software industry as a large number of real world software and web applications continue to fall short of accessibility requirements. A persisting skills deficit within the existing technology workforce has been an enduring impediment, hindering organizations from delivering truly accessible software products. This, in turn, elevates the risk of isolating and excluding a substantial portion of potential users. In this paper, we report lessons learned from a training program for teaching digital accessibility using the Communities of Practice (CoP) framework to industry professionals. We recruited 66 participants from a large multi-national software company and assigned them to two groups: one participating in a CoP and the other using self-paced learning. We report experiences from designing the training program, conducting the actual training, and assessing the efficiency of the two approaches. Based on these findings, we provide recommendations for practitioners in Learning and Development teams and educators in designing accessibility courses for industry professionals.

2401.00451

Swaroop Joshi

Parthasarathy P D and Swaroop Joshi

Exploring the Need of Accessibility Education in the Software Industry: Insights from a Survey of Software Professionals in India

To be published in International Conference on Software Engineering (ICSE'24), Software Engineering Education and Training Track

10.1145/3639474.3640079

cs.SE cs.CY

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

A UserWay study in 2021 indicates that an annual global e-commerce revenue loss of approximately $16 billion can be attributed to inaccessible websites and applications. According to the 2023 WebAIM study, only 3.7% of the world's top one million website homepages are fully accessible. This shows that many software developers use poor coding practices that don't adhere to the Web Content Accessibility Guidelines (WCAG). This research centers on software professionals and their role in addressing accessibility. This work seeks to understand (a) who within the software development community actively practices accessibility, (b) when and how accessibility is considered in the software development lifecycle, (c) the various challenges encountered in building accessible software, and (d) the resources required by software professionals to enhance product accessibility. Our survey of 269 software professionals from India sheds light on the pressing need for accessibility education within the software industry. A substantial majority (69.9%, N=269) of respondents express the need for training materials, workshops, and bootcamps to enhance their accessibility skills. We present a list of actionable recommendations that can be implemented within the industry to promote accessibility awareness and skills. We also open source our raw data for further research, encouraging continued exploration in this domain.

2401.00703

Rajes Ghosh

Chanchal Sharma, Rajes Ghosh and Sudipta Sarkar

Exploring Ladder Symmetry and Love Numbers for Static and Rotating Black Holes

Title changed, References added, Matches with the version to be published in Physical Review D as a Letter

Phys.Rev.D 109 (2024) 4, L041505

10.1103/PhysRevD.109.L041505

gr-qc

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Black hole solutions of general relativity exhibit a symmetry for the static perturbations around these spacetimes, known as "ladder symmetry". This symmetry proves useful in constructing a tower of solutions for perturbations and elucidating their general properties. Specifically, the presence of this symmetry leads to vanishing of the tidal love number associated with black holes. In this work, we find the most general spherical symmetric and static black hole spacetime that accommodates this ladder symmetry for scalar perturbation. Furthermore, we extend our calculations beyond spherical symmetry to find the class of stationary Konoplya-Rezzolla-Zhidenko black holes, which also possess a similar ladder structure.

2401.00890

Gideon Vos

Souvik Banerjee, Gideon Vos

Behind-the-horizon excitations from a single 2d CFT

25 pages, 4 figures, 2 appendices

hep-th

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

In this work, we consider the atypical non-equilibrium state found in [1708.06328] which holographically represents a behind-the-horizon excitation in a blackhole spacetime. The special feature of this state is that it looks like an equilibrium state when probed by a class of low-energy operators. First, we retrieve this property using the uniformization mapping in the limit of large central charge, in the process we are able to derive rather than presume approximate thermal physics. Furthermore, in the large-c and high-energy limit we realize these excitations as elements of the commutant algebra of a GNS-representation of the light operator algebra. Instead of analytically continuing a mixed heavy-light Euclidean correlator to a Lorentzian correlator, we identify the Euclidean correlator as a GNS-linear form and interpret the Lorentzian correlator as a vacuum expectation value of representatives of the light operator algebra on the GNS-vacuum.

2401.01159

Aneta Wojnar

Aneta Wojnar

Bose-Einstein Condensate and Liquid Helium He$^4$: Implications of GUP and Modified Gravity Correspondence

9 pages, 3 figures

Physical Review D (Vol. 109, No. 12) 2024

10.1103/PhysRevD.109.124031

gr-qc hep-ph

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Utilizing the recently established connection between Palatini-like gravity and linear Generalized Uncertainty Principle (GUP) models, we have formulated an approach that facilitates the examination of Bose gases. Our primary focus is on the ideal Bose-Einstein condensate and liquid helium, chosen as illustrative examples to underscore the feasibility of tabletop experiments in assessing gravity models. The non-interacting Bose-Einstein condensate imposes constraints on linear GUP and Palatini $f(R)$ gravity (Eddington-inspired Born-Infeld gravity) within the ranges of $-10^{12}\lesssim\sigma\lesssim 3\times 10^{24}{\text{ s}}/{\text{kg m}}$ and $-10^{-1}\lesssim\bar\beta\lesssim 10^{11} \text{ m}^2$ ($-4\times10^{-1}\lesssim\epsilon\lesssim 4\times 10^{11} \text{ m}^2$), respectively. In contrast, the properties of liquid helium suggest more realistic bounds, specifically $-10^{23}\lesssim\sigma\lesssim 10^{23}{\text{ s}}/{\text{kg m}}$ and $-10^{9}\lesssim\bar\beta\lesssim 10^{9} \text{ m}^2$. Additionally, we argue that the newly developed method employing Earth seismic waves provides improved constraints for quantum and modified gravity by approximately one order of magnitude.

2401.02104

Z. H. Wang

Wei Zhao, Tian Tian, Zhihai Wang

Single-photon scattering and bound states in a one-dimensional waveguide with topological giant atom

7 Pages, 6 Figures, Comments are welcomed

quant-ph

http://creativecommons.org/publicdomain/zero/1.0/

We investigate the single photon scattering and bound states in a coupled resonator waveguide (CRW) which couples to a topological giant atom (TGA) via two distant sites. Here, the TGA is constructed by a one dimensional Su-Schrieffer-Heeger (SSH) chain with finite length. By modulating the topological phase of the TGA, the incident photon in the CRW can be completely reflected or transmitted, and is therefore beneficial to design the coherent photonic device. Meanwhile, we also achieve two pairs of bound states locating respectively above and blow the continuum. Whether the gap is open or closed depends on the boundary condition of the TGA. Therefore, the combination of the topology and the interference provides us an exciting opportunity to manipulate the photonic state in the context of waveguide quantum electrodynamics.

2401.02216

Donghwan Lee

Donghwan Lee and Do-Wan Kim

Harnessing Membership Function Dynamics for Stability Analysis of T-S Fuzzy Systems

arXiv admin note: substantial text overlap with arXiv:2309.06841

eess.SY cs.SY math.OC

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

The main goal of this paper is to develop a new linear matrix inequality (LMI) condition for the asymptotic stability of continuous-time Takagi-Sugeno (T-S) fuzzy systems. A key advantage of this new condition is its independence from the bounds on the time-derivatives of the membership functions, a requirement present in the existing approaches. This is achieved by introducing a novel fuzzy Lyapunov function that incorporates an augmented state vector. Notably, this augmented state vector encompasses the membership functions, allowing the dynamics of these functions to be integrated into the proposed condition. This inclusion of additional information about the membership function serves to reduce the conservativeness of the suggested stability condition. To demonstrate the effectiveness of the proposed method, examples are provided.

2401.02454

Longlong Wu

Longlong Wu, Shinjae Yoo, Yong S. Chu, Xiaojing Huang, and Ian K. Robinson

Dose-efficient Automatic Differentiation for Ptychographic Reconstruction

26 pages, 5 figures

Optica 11, 821-830 (2024)

10.1364/OPTICA.522380

physics.med-ph physics.optics

http://creativecommons.org/licenses/by/4.0/

Ptychography, as a powerful lensless imaging method, has become a popular member of the coherent diffractive imaging family over decades of development. The ability to utilize low-dose X-rays and/or fast scans offers a big advantage in a ptychographic measurement (for example, when measuring radiation-sensitive samples), but results in low-photon statistics, making the subsequent phase retrieval challenging. Here, we demonstrate a dose-efficient automatic differentiation framework for ptychographic reconstruction (DAP) at low-photon statistics and low overlap ratio. As no reciprocal space constraint is required in this DAP framework, the framework, based on various forward models, shows superior performance under these conditions. It effectively suppresses potential artifacts in the reconstructed images, especially for the inherent periodic artifact in a raster scan. We validate the effectiveness and robustness of this method using both simulated and measured datasets.

2401.03160

Zilin Huang

Zilin Huang, Zihao Sheng, Chengyuan Ma, Sikai Chen

HAIM-DRL: Enhanced Human-in-the-loop Reinforcement Learning for Safe and Efficient Autonomous Driving

Accepted by Communications in Transportation Research

10.1016/j.commtr.2024.100127

cs.LG cs.AI cs.RO

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Despite significant progress in autonomous vehicles (AVs), the development of driving policies that ensure both the safety of AVs and traffic flow efficiency has not yet been fully explored. In this paper, we propose an enhanced human-in-the-loop reinforcement learning method, termed the Human as AI mentor-based deep reinforcement learning (HAIM-DRL) framework, which facilitates safe and efficient autonomous driving in mixed traffic platoon. Drawing inspiration from the human learning process, we first introduce an innovative learning paradigm that effectively injects human intelligence into AI, termed Human as AI mentor (HAIM). In this paradigm, the human expert serves as a mentor to the AI agent. While allowing the agent to sufficiently explore uncertain environments, the human expert can take control in dangerous situations and demonstrate correct actions to avoid potential accidents. On the other hand, the agent could be guided to minimize traffic flow disturbance, thereby optimizing traffic flow efficiency. In detail, HAIM-DRL leverages data collected from free exploration and partial human demonstrations as its two training sources. Remarkably, we circumvent the intricate process of manually designing reward functions; instead, we directly derive proxy state-action values from partial human demonstrations to guide the agents' policy learning. Additionally, we employ a minimal intervention technique to reduce the human mentor's cognitive load. Comparative results show that HAIM-DRL outperforms traditional methods in driving safety, sampling efficiency, mitigation of traffic flow disturbance, and generalizability to unseen traffic scenarios. The code and demo videos for this paper can be accessed at: https://zilin-huang.github.io/HAIM-DRL-website/

2401.03187

Shi-Jie Ma

Shi-Jie Ma, Rui-Bo Wang, Jian-Bo Deng, Xian-Ru Hu

Euler-Heisenberg black hole surrounded by perfect fluid dark matter

34 pages, 11 figures

10.1140/epjc/s10052-024-12914-3

gr-qc

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

A generation method of new metric in the case of static spherically symmetric space-time is derived. Using this approach, we construct a metric which describes Euler-Heisenberg black hole surrounded by perfect fluid dark matter and investigate its optical and thermodynamic properties. We found that radius of shadow will increase with the increase of dark matter effect, and more strong dark matter will diminish the light intensity of accretion disk generally. Moreover, in thermodynamics, when quantum electrodynamic parameter is positive, there will be a critical value of dark matter parameter, which determine the number of black hole's critical points.

2401.04056

Gokul Swamy

Gokul Swamy, Christoph Dann, Rahul Kidambi, Zhiwei Steven Wu, Alekh Agarwal

A Minimaximalist Approach to Reinforcement Learning from Human Feedback

cs.LG

http://creativecommons.org/licenses/by-sa/4.0/

We present Self-Play Preference Optimization (SPO), an algorithm for reinforcement learning from human feedback. Our approach is minimalist in that it does not require training a reward model nor unstable adversarial training and is therefore rather simple to implement. Our approach is maximalist in that it provably handles non-Markovian, intransitive, and stochastic preferences while being robust to the compounding errors that plague offline approaches to sequential prediction. To achieve the preceding qualities, we build upon the concept of a Minimax Winner (MW), a notion of preference aggregation from the social choice theory literature that frames learning from preferences as a zero-sum game between two policies. By leveraging the symmetry of this game, we prove that rather than using the traditional technique of dueling two policies to compute the MW, we can simply have a single agent play against itself while maintaining strong convergence guarantees. Practically, this corresponds to sampling multiple trajectories from a policy, asking a preference or teacher model to compare them, and then using the proportion of wins as the reward for a particular trajectory. We demonstrate that on a suite of continuous control tasks, we are able to learn significantly more efficiently than reward-model based approaches while maintaining robustness to the intransitive and stochastic preferences that frequently occur in practice when aggregating human judgments.

2401.04206

Robert Kaufman

Robert Kaufman, Jean Costa, Everlyne Kimani

Effects of Multimodal Explanations for Autonomous Driving on Driving Performance, Cognitive Load, Expertise, Confidence, and Trust

14 pages, published in Scientific Reports

Scientific Reports volume 14, Article number: 13061 (2024)

10.1038/s41598-024-62052-9

cs.HC cs.AI cs.RO

http://creativecommons.org/licenses/by/4.0/

Advances in autonomous driving provide an opportunity for AI-assisted driving instruction that directly addresses the critical need for human driving improvement. How should an AI instructor convey information to promote learning? In a pre-post experiment (n = 41), we tested the impact of an AI Coach's explanatory communications modeled after performance driving expert instructions. Participants were divided into four (4) groups to assess two (2) dimensions of the AI coach's explanations: information type ('what' and 'why'-type explanations) and presentation modality (auditory and visual). We compare how different explanatory techniques impact driving performance, cognitive load, confidence, expertise, and trust via observational learning. Through interview, we delineate participant learning processes. Results show AI coaching can effectively teach performance driving skills to novices. We find the type and modality of information influences performance outcomes. Differences in how successfully participants learned are attributed to how information directs attention, mitigates uncertainty, and influences overload experienced by participants. Results suggest efficient, modality-appropriate explanations should be opted for when designing effective HMI communications that can instruct without overwhelming. Further, results support the need to align communications with human learning and cognitive processes. We provide eight design implications for future autonomous vehicle HMI and AI coach design.

2401.05332

Jim Peter Skulte

Jim Skulte, Phatthamon Kongkhambut, Hans Ke{\ss}ler, Andreas Hemmerich, Ludwig Mathey and Jayson G. Cosme

Realizing limit cycles in dissipative bosonic systems

Phys. Rev. A 109, 063317 (2024)

10.1103/PhysRevA.109.063317

cond-mat.quant-gas quant-ph

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We propose a general mechanism for generating limit cycle (LC) oscillations by coupling a linear bosonic mode to a dissipative nonlinear bosonic mode. By analyzing the stability matrix, we show that LCs arise due to a supercritical Hopf bifurcation. We find that the existence of LCs is independent of the sign of the effective nonlinear interaction. The LC phase can be classified as a continuous time crystal (CTC), if it emerges in a many-body system. The bosonic model can be realised in three-level systems interacting with a quantised light mode as realised in atom-cavity systems. Using such a platform, we experimentally observe LCs for the first time in an atom-cavity system with attractive optical pump lattice, thereby confirming our theoretical predictions.

2401.06112

Hiroaki Yamagiwa

Hiroaki Yamagiwa, Yusuke Takase, Hidetoshi Shimodaira

Axis Tour: Word Tour Determines the Order of Axes in ICA-transformed Embeddings

cs.CL

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Word embedding is one of the most important components in natural language processing, but interpreting high-dimensional embeddings remains a challenging problem. To address this problem, Independent Component Analysis (ICA) is identified as an effective solution. ICA-transformed word embeddings reveal interpretable semantic axes; however, the order of these axes are arbitrary. In this study, we focus on this property and propose a novel method, Axis Tour, which optimizes the order of the axes. Inspired by Word Tour, a one-dimensional word embedding method, we aim to improve the clarity of the word embedding space by maximizing the semantic continuity of the axes. Furthermore, we show through experiments on downstream tasks that Axis Tour yields better or comparable low-dimensional embeddings compared to both PCA and ICA.

2401.07555

Marta Zoppello

Cristina Giannotti, Andrea Spiro and Marta Zoppello

Distributions and controllability problems (I)

50 pages, 3 figures This paper is a revision of the previous manuscript, now we removed an unnecessary assumption, enhancing in this way all results, which now hold in a much more general setting

math.OC math.DG

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We consider a non-linear real analytic control system of first order $\dot q^i = f^i(t, q, w)$, with controls $w = (w^\alpha)$ in a connected open set $\mathcal{K} \subset \mathbb{R}^m$ and configurations $q = (q^i)$ in $\mathcal{Q} := \mathbb{R}^n$. The set of points in the extended space-time $\mathcal{M} = \mathbb{R} \times \mathcal{Q} \times \mathcal{K}$, which can be reached from a triple $x_o = (t_o , q_o, w_o) \in \mathcal{M}$ through a continuous graph completion $\gamma(s) = \big(t(s), q(t(s)), w(t(s))\big)$ of the graph of a solution $t \to (q(t), w(t))$, $t \in [t_o ,t_o + T]$, with piecewise real analytic controls, is called the {\it $\mathcal{M}$-attainable set of $x_o$ in time $T$}. We prove that if $y_o$ is an $\mathcal{M}$-attainable point of $x_o$, a large set of other nearby $\mathcal{M}$-attainable points of $x_o$ can be determined starting directly from $y_o$ and applying an appropriate ordered composition of flows of vector fields in a distinguished distribution $\mathcal{D}^{II} \subset T \mathcal{M}$, canonically associated with the control system. We then determine sufficient conditions for such neighbouring points to constitute an orbit of the pseudogroup of local diffeomorphisms generated by the vector fields in $\mathcal{D}^{II}$. If such conditions are satisfied and if the tangent spaces of these orbits have maximal rank projections onto $\mathcal{Q}$, the control system is locally accessible and has the small time local controllability property near the state points of equilibrium. These results lead to new proofs of classical local controllability criterions and yield new methods to establish the accessibility and the small time local controllability of non-linear control systems.

2401.07560

Marta Zoppello

Cristina Giannotti, Andrea Spiro and Marta Zoppello

Distributions and controllability problems (II)

23 pages, 2 figures This paper is a revision of the previous manuscript, now we removed an unnecessary assumption, enhancing in this way all results, which now hold in a much more general setting

math.OC

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

In [C. Giannotti, A. Spiro, M. Zoppello, {\it Distributions and controllability problems (I)}, preprint posted on ArXiv (2024)], we introduced a new approach to the real analytic non-linear control systems of the form $\dot q^i = f^i(t, q, w)$, with controls $w = (w^\alpha)$ running in a connected open set $\mathcal{K}$ of $ \mathbb{R}^m$ and states represented by points $q = (q^i)$ in a configuration space $\mathcal{Q} := \mathbb{R}^n$. The new approach consists of a differential-geometric study of (a) the oriented piecewise regular curves in the {\it extended space-time} $\mathcal{M} = \mathbb{R} \times \mathcal{Q} \times \mathcal{K}$, which are the (completed) graphs of the piecewise real analytic solutions $t \mapsto (q(t), w(t))$ of the control system, and (b) the local structure of the sets of points of $\mathcal{M}$ that are reachable from an initial point $x_o = (t_o, q_o, w_o) \in \mathcal{M}$ through such (completed) graphs. The main results of that paper are two new criterions which can be used to establish the small time local controllability near stable points of real analytic non-linear systems. The goal of this paper is to offer a friendly user's guide to those criterions, illustrating them by several examples. In particular, we analyse certain non-linear control systems, for which the new criterions show that they are small time locally controllable at their stable points, while, at the best of our knowledge, all other previous criterions are either inconclusive or not applicable.

2401.07573

A Daassou

Hayat. Laassiri, Ahmed. Daassou, Rachid. Benbrik

Thermodynamic features of AdS black holes within the Rastall gravity and perfect fluid matter framework

16 pages, 3 figures. Accepted for publication in Physica Scripta (2024)

Phys. Scr. 99 075024 (2024)

10.1088/1402-4896/ad52d1

hep-th

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

In this study, we analyzed the impact of a perfect fluid on the phase transition of Anti-de Sitter (AdS) black holes within the Rastall gravitational background. Compared to similar studies in the literature, the findings of this work are highlighted by the determination of analytical expressions for critical points for charged and Kerr-Newman AdS black holes using approximated formulas for the horizon radius. An accurate analysis of these new analytical expressions allowed us to discover a new viable condition that relates the Rastall parameter \k{appa}{\lambda} to the equation of state parameter {\omega}, expressed as: \k{appa}{\lambda} = {\omega}/(1+{\omega}). Thanks to this new condition, we were able to reproduce all the analytical expressions of critical points calculated within the framework of Einstein\^as general relativity for two cases: a charged AdS black hole and a rotating AdS black hole. These findings suggest that Rastall gravity, considering this new condition, could serve as an alternative theory of gravitation to general relativity. The approximate expression of the horizon radius also enabled the exploration of the distinctiveness of fractional-order phase transitions in these AdS black holes. Furthermore, we calculated the critical exponents, offering insights into the behavior of crucial thermodynamic quantities near the inflection point. Examining how a perfect fluid influences phase transition reveals various critical behaviors, demonstrating that the variation in the phase transition depends on the intensity of the perfect fluid. Notably, this variation is portrayed by a linearly increasing trajectory with the escalation of this intensity.

2401.07577

Jes\'us Garc\'ia D\'iaz

Jes\'us Garc\'ia-D\'iaz and Jos\'e Alejandro Cornejo-Acosta

A greedy heuristic for graph burning

15 pages

cs.DM math.CO

http://creativecommons.org/licenses/by-nc-nd/4.0/

Given a graph $G$, the optimization version of the graph burning problem seeks for a sequence of vertices, $(u_1,u_2,...,u_k) \in V(G)^k$, with minimum $k$ and such that every $v \in V(G)$ has distance at most $k-i$ to some vertex $u_i$. The length $k$ of the optimal solution is known as the burning number and is denoted by $b(G)$, an invariant that helps quantify the graph's vulnerability to contagion. This paper explores the advantages and limitations of an $\mathcal{O}(mn + kn^2)$ deterministic greedy heuristic for this problem, where $n$ is the graph's order, $m$ is the graph's size, and $k$ is a guess on $b(G)$. This heuristic is based on the relationship between the graph burning problem and the clustered maximum coverage problem, and despite having limitations on paths and cycles, it found most of the optimal and best-known solutions of benchmark and synthetic graphs with up to 102400 vertices.

2401.07759

Peter Gothen

Pedro M. Silva and Peter B. Gothen

The conformal limit and projective structures

24 pages, comments welcome. V2: minor corrections and improvements. To appear in IMRN

10.1093/imrn/rnae142

math.DG math.AG

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

The non-abelian Hodge correspondence maps a polystable $\mathrm{SL}(2,\mathbb{R})$-Higgs bundle on a compact Riemann surface $X$ of genus $g\geq2$ to a connection which, in some cases, is the holonomy of a branched hyperbolic structure. On the other hand, Gaiotto's conformal limit maps the same bundle to a partial oper, i.e., to a connection whose holonomy is that of a branched complex projective structure compatible with $X$. In this article, we show how these are both instances of the same phenomenon: the family of connections appearing in the conformal limit can be understood as a family of complex projective structures, deforming the hyperbolic ones into the ones compatible with $X$. We also show that, when the Higgs bundle has zero Toledo invariant, this deformation is optimal, inducing a geodesic on Teichm\"uller's metric space.

2401.07912

Jordi Rudo Weggemans

Jordi Weggemans

Lower Bounds for Unitary Property Testing with Proofs and Advice

26 pages, 2 figures, V2: writing improvements and minor fixes, fixed a bug in the oracle separation (at the cost of having slightly weaker separations now)

quant-ph cs.CC

http://creativecommons.org/licenses/by/4.0/

In unitary property testing a quantum algorithm, also known as a tester, is given query access to a black-box unitary and has to decide whether it satisfies some property. We propose a new technique for proving lower bounds on the quantum query complexity of unitary property testing and related problems, which utilises its connection to unitary channel discrimination. The main advantage of this technique is that all obtained lower bounds hold for any $\mathsf{C}$-tester with $\mathsf{C} \subseteq \mathsf{QMA}(2)/\mathsf{qpoly}$, showing that even having access to both (unentangled) quantum proofs and advice does not help for many unitary problems. We apply our technique to prove lower bounds for problems like quantum phase estimation, the entanglement entropy problem, quantum Gibbs sampling and more, removing all logarithmic factors in the lower bounds obtained by the sample-to-query lifting theorem of Wang and Zhang (2023). As a direct corollary, we show that there exist quantum oracles relative to which $\mathsf{QMA}(2) \not\supset \mathsf{SBQP}$ and $\mathsf{QMA}/\mathsf{qpoly} \not\supset \mathsf{SBQP}$. The former shows that, at least in a black-box way, having unentangled quantum proofs does not help in solving problems that require high precision.

2401.08316

M Said Ergoktas

M. Said Ergoktas, Ali Kecebas, Konstantinos Despotelis, Sina Soleymani, Gokhan Bakan, Askin Kocabas, Alessandro Principi, Stefan Rotter, Sahin K. Ozdemir, Coskun Kocabas

Localised Thermal Emission from Topological Interfaces

10 pages, 4 figures

Science 384, 1122-1126 (2024)

10.1126/science.ado0534

physics.app-ph cond-mat.mtrl-sci physics.optics

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

The control of thermal radiation by shaping its spatial and spectral emission characteristics plays a key role in many areas of science and engineering. Conventional approaches to tailor thermal emission using metamaterials are severely hampered both by the limited spatial resolution of the required sub-wavelength material structures and by the materials' strong absorption in the infrared. Here, we demonstrate a promising new approach based on the concept of topology. By changing a single parameter of a multilayer coating, we control the reflection topology of a surface, with the critical point of zero reflection being topologically protected. As a result, the boundaries between sub-critical and super-critical spatial domains host topological interface states with near-unity thermal emissivity. Our experimental demonstration of this effect shows that topological concepts enable unconventional manipulation of thermal light with promising applications for thermal management, energy harvesting and thermal camouflage.

2401.08397

Enrico Magliano

Enrico Magliano, Alessio Carpegna, Alessadro Savino, Stefano Di Carlo

A Micro Architectural Events Aware Real-Time Embedded System Fault Injector

2024 IEEE 25th Latin American Test Symposium (LATS)

10.1109/LATS62223.2024.10534595

cs.AR cs.AI

http://creativecommons.org/licenses/by-nc-sa/4.0/

In contemporary times, the increasing complexity of the system poses significant challenges to the reliability, trustworthiness, and security of the SACRES. Key issues include the susceptibility to phenomena such as instantaneous voltage spikes, electromagnetic interference, neutron strikes, and out-of-range temperatures. These factors can induce switch state changes in transistors, resulting in bit-flipping, soft errors, and transient corruption of stored data in memory. The occurrence of soft errors, in turn, may lead to system faults that can propel the system into a hazardous state. Particularly in critical sectors like automotive, avionics, or aerospace, such malfunctions can have real-world implications, potentially causing harm to individuals. This paper introduces a novel fault injector designed to facilitate the monitoring, aggregation, and examination of micro-architectural events. This is achieved by harnessing the microprocessor's PMU and the debugging interface, specifically focusing on ensuring the repeatability of fault injections. The fault injection methodology targets bit-flipping within the memory system, affecting CPU registers and RAM. The outcomes of these fault injections enable a thorough analysis of the impact of soft errors and establish a robust correlation between the identified faults and the essential timing predictability demanded by SACRES.

2401.08923

Aydogan Ozcan

Jingtian Hu, Kun Liao, Niyazi Ulas Dinc, Carlo Gigli, Bijie Bai, Tianyi Gan, Xurong Li, Hanlong Chen, Xilin Yang, Yuhang Li, Cagatay Isil, Md Sadman Sakib Rahman, Jingxi Li, Xiaoyong Hu, Mona Jarrahi, Demetri Psaltis, and Aydogan Ozcan

Subwavelength Imaging using a Solid-Immersion Diffractive Optical Processor

32 Pages, 9 Figures

eLight (2024)

10.1186/s43593-024-00067-5

physics.optics cs.CV physics.app-ph

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Phase imaging is widely used in biomedical imaging, sensing, and material characterization, among other fields. However, direct imaging of phase objects with subwavelength resolution remains a challenge. Here, we demonstrate subwavelength imaging of phase and amplitude objects based on all-optical diffractive encoding and decoding. To resolve subwavelength features of an object, the diffractive imager uses a thin, high-index solid-immersion layer to transmit high-frequency information of the object to a spatially-optimized diffractive encoder, which converts/encodes high-frequency information of the input into low-frequency spatial modes for transmission through air. The subsequent diffractive decoder layers (in air) are jointly designed with the encoder using deep-learning-based optimization, and communicate with the encoder layer to create magnified images of input objects at its output, revealing subwavelength features that would otherwise be washed away due to diffraction limit. We demonstrate that this all-optical collaboration between a diffractive solid-immersion encoder and the following decoder layers in air can resolve subwavelength phase and amplitude features of input objects in a highly compact design. To experimentally demonstrate its proof-of-concept, we used terahertz radiation and developed a fabrication method for creating monolithic multi-layer diffractive processors. Through these monolithically fabricated diffractive encoder-decoder pairs, we demonstrated phase-to-intensity transformations and all-optically reconstructed subwavelength phase features of input objects by directly transforming them into magnified intensity features at the output. This solid-immersion-based diffractive imager, with its compact and cost-effective design, can find wide-ranging applications in bioimaging, endoscopy, sensing and materials characterization.

2401.09050

Zike Wu

Zike Wu, Pan Zhou, Xuanyu Yi, Xiaoding Yuan, Hanwang Zhang

Consistent3D: Towards Consistent High-Fidelity Text-to-3D Generation with Deterministic Sampling Prior

Accepted to CVPR 2024

cs.CV cs.LG

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Score distillation sampling (SDS) and its variants have greatly boosted the development of text-to-3D generation, but are vulnerable to geometry collapse and poor textures yet. To solve this issue, we first deeply analyze the SDS and find that its distillation sampling process indeed corresponds to the trajectory sampling of a stochastic differential equation (SDE): SDS samples along an SDE trajectory to yield a less noisy sample which then serves as a guidance to optimize a 3D model. However, the randomness in SDE sampling often leads to a diverse and unpredictable sample which is not always less noisy, and thus is not a consistently correct guidance, explaining the vulnerability of SDS. Since for any SDE, there always exists an ordinary differential equation (ODE) whose trajectory sampling can deterministically and consistently converge to the desired target point as the SDE, we propose a novel and effective "Consistent3D" method that explores the ODE deterministic sampling prior for text-to-3D generation. Specifically, at each training iteration, given a rendered image by a 3D model, we first estimate its desired 3D score function by a pre-trained 2D diffusion model, and build an ODE for trajectory sampling. Next, we design a consistency distillation sampling loss which samples along the ODE trajectory to generate two adjacent samples and uses the less noisy sample to guide another more noisy one for distilling the deterministic prior into the 3D model. Experimental results show the efficacy of our Consistent3D in generating high-fidelity and diverse 3D objects and large-scale scenes, as shown in Fig. 1. The codes are available at https://github.com/sail-sg/Consistent3D.

2401.09123

Kay Joerg Wiese

Kay Joerg Wiese

Hyperuniformity in the Manna Model, Conserved Directed Percolation and Depinning

5 pages, 3 figures. v3 contains additional material

cond-mat.stat-mech cond-mat.dis-nn

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Hyperuniformity is an emergent property, whereby the structure factor of the density $n$ scales as $S(q) \sim q^\alpha$, with $\alpha>0$. We show that for the conserved directed percolation (CDP) class, to which the Manna model belongs, there is an exact mapping between the density $n$ in CDP, and the interface position $u$ at depinning, $n(x)=n_0+\nabla^2 u(x)$, where $n_0$ is the conserved particle density. As a consequence, the hyperuniformity exponent equals $\alpha={4-d-2\zeta}$, with $\zeta$ the roughness exponent at depinning, and $d$ the dimension. In $d=1$, $\alpha=1/2$, while $0.6> \alpha\ge0$ for other $d$. Our results fit well simulations in the literature, except in $d=1$, where we perform our own to confirm this result. Such an exact relation between two seemingly different fields is surprising, and paves new paths to think about hyperuniformity and depinning. As corollaries, we get results of unprecedented precision in all dimensions, exact in $d=1$. This corrects earlier work on hyperuniformity in CDP.

2401.09187

Mahdi Hajlaoui

M. Hajlaoui, S.W. D'Souza, L. \v{S}mejkal, D. Kriegner, G. Krizman, T. Zakusylo, N. Olszowska, O. Caha, J. Michali\v{c}ka, A. Marmodoro, K. V\'yborn\'y, A. Ernst, M. Cinchetti, J. Minar, T. Jungwirth, and G. Springholz

Temperature dependence of relativistic valence band splitting induced by an altermagnetic phase transition

10.1002/adma.202314076

cond-mat.mtrl-sci

http://creativecommons.org/licenses/by/4.0/

Altermagnetic (AM) materials exhibit non-relativistic, momentum-dependent spin-split states, ushering in new opportunities for spin electronic devices. While the characteristics of spin-splitting have been documented within the framework of the non-relativistic spin group symmetry, there has been limited exploration of the inclusion of relativistic symmetry and its impact on the emergence of a novel spin-splitting in the band structure. This study delves into the intricate relativistic electronic structure of an AM material, alpha-MnTe. Employing temperature-dependent angle-resolved photoelectron spectroscopy across the AM phase transition, we elucidate the emergence of a relativistic valence band splitting concurrent with the establishment of magnetic order. This discovery is validated through disordered local moment calculations, modeling the influence of magnetic order on the electronic structure and confirming the magnetic origin of the observed splitting. The temperature-dependent splitting is ascribed to the advent of relativistic spin-splitting resulting from the strengthening of AM order in alpha-MnTe as the temperature decreases. This sheds light on a previously unexplored facet of this intriguing material.

2401.09907

Liang-Bi Wu

Li-Ming Cao, Jia-Ning Chen, Liang-Bi Wu, Libo Xie and Yu-Sen Zhou

The pseudospectrum and spectrum (in)stability of quantum corrected Schwarzschild black hole

23 pages, 10 figures. This manuscript has been accepted for publication as a regular article in SCPMA

ICTS-USTC/PCFT-24-03

gr-qc

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

In this study, we investigate the pseudospectrum and spectrum (in)stability of quantum corrected Schwarzschild black hole. Methodologically, we use the hyperboloidal framework to cast the quasinormal mode (QNM) problem into an eigenvalue problem associated with a non-selfadjoint operator, and then the spectrum and pseudospectrum are depicted. Besides, the invariant subspace method is exploited to improve the computational efficiency for pseudospectrum. The investigation into the spectrum (in)stability entails two main aspects. On the one hand, we calculate the spectra of the quantum corrected black hole, then by the means of the migration ratio, the impact of the quantum correction effect on the Schwarzschild black hole has been studied. The results indicate that the so-called ``migration ratio instability" will occur for small black holes with small angular momentum number l. In the eikonal limit, the migration ratios remain the same for each overtone. On the other hand, we study the spectrum (in)stability of the quantum corrected black hole by directly adding some particular perturbations into the effective potential, where perturbations are located at the event horizon and null infinity, respectively. There are two interesting observations under the same perturbation energy norm. First, perturbations at infinity are more capable of generating spectrum instability than those at the event horizon. Second, we find that the peak distribution can lead to the instability of QNM spectrum more efficiently than the average distribution.

2401.10149

Alec Wilson

Alec Wilson, Ryan Menzies, Neela Morarji, David Foster, Marco Casassa Mont, Esin Turkbeyler, Lisa Gralewski

Multi-Agent Reinforcement Learning for Maritime Operational Technology Cyber Security

13 pages, 7 figures, Proceedings of the Conference on Applied Machine Learning in Information Security 2023 (CAMLIS)

Proceedings of the Conference on Applied Machine Learning in Information Security 2023 (CAMLIS), Arlington VA, USA, October 19-20, 2023, CEUR-WS.org, online CEUR-WS.org/Vol-3652/paper3.pdf

cs.LG cs.CR cs.MA

http://creativecommons.org/licenses/by/4.0/

This paper demonstrates the potential for autonomous cyber defence to be applied on industrial control systems and provides a baseline environment to further explore Multi-Agent Reinforcement Learning's (MARL) application to this problem domain. It introduces a simulation environment, IPMSRL, of a generic Integrated Platform Management System (IPMS) and explores the use of MARL for autonomous cyber defence decision-making on generic maritime based IPMS Operational Technology (OT). OT cyber defensive actions are less mature than they are for Enterprise IT. This is due to the relatively brittle nature of OT infrastructure originating from the use of legacy systems, design-time engineering assumptions, and lack of full-scale modern security controls. There are many obstacles to be tackled across the cyber landscape due to continually increasing cyber-attack sophistication and the limitations of traditional IT-centric cyber defence solutions. Traditional IT controls are rarely deployed on OT infrastructure, and where they are, some threats aren't fully addressed. In our experiments, a shared critic implementation of Multi Agent Proximal Policy Optimisation (MAPPO) outperformed Independent Proximal Policy Optimisation (IPPO). MAPPO reached an optimal policy (episode outcome mean of 1) after 800K timesteps, whereas IPPO was only able to reach an episode outcome mean of 0.966 after one million timesteps. Hyperparameter tuning greatly improved training performance. Across one million timesteps the tuned hyperparameters reached an optimal policy whereas the default hyperparameters only managed to win sporadically, with most simulations resulting in a draw. We tested a real-world constraint, attack detection alert success, and found that when alert success probability is reduced to 0.75 or 0.9, the MARL defenders were still able to win in over 97.5% or 99.5% of episodes, respectively.

2401.10645

Shuang Li

Kewei Yu, Jiazhen Peng, Shuang Li, Kejun Wu, Wei Xie and Fei Sun

Photoproduction of lepton pair in ultra-relativistic heavy-ion collisions

10 pages, 9 figures

Phys. Rev. C 109, 064907 (2024)

10.1103/PhysRevC.109.064907

nucl-th hep-ph

http://creativecommons.org/licenses/by/4.0/

Dilepton production provides a unique probe of the strong electromagnetic field produced in heavy-ion collisions. To map out the behavior of its transverse momentum broadening, we present a theoretical model based on the equivalent photon approximation, and then we update it to make direct comparisons with the recent experimental measurements. We find that the model calculations can describe well, not only the average transverse momentum squared of $e^{+}e^{-}$ pairs in Au--Au collisions at $\sqrt{s_{\rm NN}}=200$ GeV, but also the acoplanarity of $\mu^{+}\mu^{-}$ pairs in Pb--Pb collisions at$\sqrt{s_{\rm NN}}=5.02$ TeV. Furthermore, the model predictions are also able to reproduce the measured dependencies of the pair mass and the transverse momentum squared.

2401.10982

Nicholas Fazio

Nicholas Fazio, Robin Harper and Stephen Bartlett

Logical Noise Bias in Magic State Injection

11+5 pages, 17 figures. Comments welcome

quant-ph

http://creativecommons.org/licenses/by-sa/4.0/

Fault-tolerant architectures aim to reduce the noise of a quantum computation. Despite such architectures being well studied a detailed understanding of how noise is transformed in a fault-tolerant primitive such as magic state injection is currently lacking. We use numerical simulations of logical process tomography on a fault-tolerant gadget that implements a logical T = $Z({\pi}/8)$ gate using magic state injection, to understand how noise characteristics at the physical level are transformed into noise characteristics at the logical level. We show how, in this gadget, a significant phase ($Z$) bias can arise in the logical noise, even with unbiased noise at the physical level. While the magic state injection gadget intrinsically induces biased noise, with extant phase bias being further amplified at the logical level, we identify noisy error correction circuits as a key limiting factor on the magnitude of this logical noise bias. Our approach provides a framework for assessing the detailed noise characteristics, as well as the overall performance, of fault-tolerant logical primitives.

2401.11422

Haruki Kono

Haruki Kono

Local Identification in Instrumental Variable Multivariate Quantile Regression Models

econ.EM math.ST stat.TH

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

In the instrumental variable quantile regression (IVQR) model of Chernozhukov and Hansen (2005), a one-dimensional unobserved rank variable monotonically determines a single potential outcome. Even when multiple outcomes are simultaneously of interest, it is common to apply the IVQR model to each of them separately. This practice implicitly assumes that the rank variable of each regression model affects only the corresponding outcome and does not affect the other outcomes. In reality, however, it is often the case that all rank variables together determine the outcomes, which leads to a systematic correlation between the outcomes. To deal with this, we propose a nonlinear IV model that allows for multivariate unobserved heterogeneity, each of which is considered as a rank variable for an observed outcome. We show that the structural function of our model is locally identified under the assumption that the IV and the treatment variable are sufficiently positively correlated.

2401.11963

Pengyi Li

Pengyi Li, Jianye Hao, Hongyao Tang, Xian Fu, Yan Zheng, Ke Tang

Bridging Evolutionary Algorithms and Reinforcement Learning: A Comprehensive Survey on Hybrid Algorithms

cs.NE cs.AI cs.LG

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Evolutionary Reinforcement Learning (ERL), which integrates Evolutionary Algorithms (EAs) and Reinforcement Learning (RL) for optimization, has demonstrated remarkable performance advancements. By fusing both approaches, ERL has emerged as a promising research direction. This survey offers a comprehensive overview of the diverse research branches in ERL. Specifically, we systematically summarize recent advancements in related algorithms and identify three primary research directions: EA-assisted Optimization of RL, RL-assisted Optimization of EA, and synergistic optimization of EA and RL. Following that, we conduct an in-depth analysis of each research direction, organizing multiple research branches. We elucidate the problems that each branch aims to tackle and how the integration of EAs and RL addresses these challenges. In conclusion, we discuss potential challenges and prospective future research directions across various research directions. To facilitate researchers in delving into ERL, we organize the algorithms and codes involved on https://github.com/yeshenpy/Awesome-Evolutionary-Reinforcement-Learning.

2401.12040

Sergey Yu. Vernov

Ekaterina O. Pozdeeva and Sergey Yu. Vernov

Construction of Chiral Cosmological Models Unifying Inflation and Primordial Black Hole Formation

6 pages

SPACE, TIME AND FUNDAMENTAL INTERACTIONS, 2024, V. 1 (46), P. 90

10.17238/issn2226-8812.2024.1.90-94

gr-qc

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We propose the method for construction of $F(R,\xi)$ gravity model, unifying inflation and primordial black hole formation. The proposed models are based on the Starobinsky $R+R^2$ inflationary model, so, the function $F(R,\xi)$ is a quadratic polynomial of the Ricci scalar $R$. We show that the potential of the corresponding two-field chiral cosmological model in the Einstein frame can be always found in terms of the elementary functions. The special choice of the function $F(R,\xi)$ allows us to get such a generalization of the hybrid inflation that can describe both inflation, and the primordial black hole formation.

2401.12637

Ylann Rouzaire

Ylann Rouzaire and Daniel JG Pearce and Ignacio Pagonabarraga and Demian Levis

Non-Reciprocal Interactions Reshape Topological Defect Annihilation

cond-mat.stat-mech cond-mat.soft nlin.AO physics.comp-ph

http://creativecommons.org/licenses/by-nc-sa/4.0/

We show how non-reciprocal ferromagnetic interactions between neighbouring planar spins in two dimensions, affect the behaviour of topological defects. Non-reciprocity is introduced by weighting the coupling strength of the two-dimensional XY model by an anisotropic kernel. As a consequence, in addition to the topological charge $q$, the actual shape of the defects becomes crucial to faithfully describe their dynamics. Non-reciprocal coupling twists the spin field, selecting specific defect shapes, dramatically altering the pair annihilation process. Defect annihilation can either be enhanced or hindered, depending on the shape of the defects concerned and the degree of non-reciprocity in the system. We introduce a continuous description -- for which the phenomenological coefficients can be explicitly written in terms of the microscopic ones -- that captures the behaviour of the lattice model.

2401.12912

Llu\"isa-Maria Mir

Alexis Men\'endez-V\'azquez, Marc Andr\'es-Carcasona, Mario Mart\'inez, Llu\"isa-Maria Mir

Searches for Compact Binary Coalescence Events Using Neural Networks in LIGO/Virgo Third Observation Period

14 pages, 7 figures

Class. Quantum Grav. 41 135018 (2024)

10.1088/1361-6382/ad4f42

gr-qc astro-ph.IM

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We present the results on the search for the coalescence of compact binary mergers using convolutional neural networks and the LIGO/Virgo data for the O3 observation period. Two-dimensional images in time and frequency are used as input. The analysis is performed in three separate mass regions covering the range for the masses in the binary system from 0.2 to 100 solar masses, excluding very asymmetric mass configurations. We explore neural networks trained with input information from pairs of interferometers or all three interferometers together, concluding that the use of the maximum information available leads to an improved performance. A scan over the O3 data set, using the convolutional neural networks, is performed with different fake rate thresholds for claiming detection of at most one event per year or at most one event per week. The latter would correspond to a loose online selection still leading to affordable fake alarm rates. The efficiency of the neutral networks to detect the O3 catalog events is discussed. In the case of a fake rate threshold of at most one event per week, the scan leads to the detection of about 50% of the O3 catalog events. Once the search is limited to the catalog events within the mass range used for neural networks training, the detection efficiency increases up to 70%. A further improvement in the search efficiency, using the same kind of algorithms, will require the implementation of new criteria for the suppression of detector glitches.

2401.13640

Valerio Belocchi

Valerio Belocchi, Maria Benedetta Barbaro, Arturo De Pace, Marco Martini

Relativistic meson-exchange currents in semi-inclusive lepton scattering

14 pages, 10 figures

10.1103/PhysRevC.109.065502

nucl-th hep-ex hep-ph

http://creativecommons.org/licenses/by/4.0/

We assess the impact of two-particle--two-hole excitations on the semi-inclusive electron scattering process (e,e'p) using a fully relativistic nuclear model calculation that precisely incorporates antisymmetrization. The calculation encompasses all contributions involving the exchange of a single pion and the excitation of a Delta resonance. Our results are compared with (e,e'p) data on carbon at kinematics where two-nucleon emission dominates. This work represents an essential step towards the microscopic computation of the two-particle--two-hole contribution to semi-inclusive neutrino reactions, crucial in the analysis of neutrino oscillation experiments.

2401.15241

Masaru Isonuma

Masaru Isonuma and Ivan Titov

Unlearning Traces the Influential Training Data of Language Models

14 pages, to appear in ACL2024 main conference (long paper)

cs.CL cs.AI

http://creativecommons.org/licenses/by/4.0/

Identifying the training datasets that influence a language model's outputs is essential for minimizing the generation of harmful content and enhancing its performance. Ideally, we can measure the influence of each dataset by removing it from training; however, it is prohibitively expensive to retrain a model multiple times. This paper presents UnTrac: unlearning traces the influence of a training dataset on the model's performance. UnTrac is extremely simple; each training dataset is unlearned by gradient ascent, and we evaluate how much the model's predictions change after unlearning. Furthermore, we propose a more scalable approach, UnTrac-Inv, which unlearns a test dataset and evaluates the unlearned model on training datasets. UnTrac-Inv resembles UnTrac, while being efficient for massive training datasets. In the experiments, we examine if our methods can assess the influence of pretraining datasets on generating toxic, biased, and untruthful content. Our methods estimate their influence much more accurately than existing methods while requiring neither excessive memory space nor multiple checkpoints.

2401.15611

Luca Moleri

Abhik Jash, Luca Moleri, Andrea Tesi, Shikma Bressler

Discharge quenching mechanism and performance of RPWELL with tunable 3D printed resistive plates, charge evacuation in semiconductive glass RPWELL and discharge quenching for Cryogenic-RWELL over a wide range of resistivity

Presented in the 3rd International Conference on Detector Stability and Aging Phenomena in Gaseous Detectors. Submitted to Nuclear Inst. and Methods in Physics Research, A --Manuscript Draft--

physics.ins-det

http://creativecommons.org/licenses/by-nc-nd/4.0/

Resistive electrodes are used in gaseous detectors to quench electrical discharges. This helps to protect delicate electrodes and readout electronics and to improve the stability of the detector operation. An RPWELL is a THGEM-based WELL detector with a resistive plate coupled to a conductive anode. Till now, the choice of the resistive plate was limited to a few materials, like LRS Glass and Semitron. These materials have fixed resistivities and, sometimes, thickness and area limitations. This restricts the potential usage of the detector to a rather small range of applications, as well as the possibility of studying in depth the physics processes governing the discharge quenching mechanism. In our present study, we used a new plastic material doped with carbon nanotubes to produce resistive plates with a commercial 3D printer. This method has the flexibility to produce samples of different thicknesses and different resistivity values. We describe here the sample production and characterize the RPWELL performance with different resistive plates. In particular we show the dependence of discharge quenching on the thickness and resistivity of the plate. The dynamics of the charge carriers in the material is proposed as an explanation for the long gain recovery time after a discharge.

2401.16334

Cihan Sabuncu

Cihan Sabuncu

Right-angled triangles with almost prime hypotenuse

15 pages. Referee comments incorporated. To appear in Research in Number Theory

math.NT

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

The sequence OEIS A281505 consists of distinct odd legs in right triangles with integer sides and prime hypotenuse. In this paper, we count the closely related quantity of even legs with almost prime hypotenuse. More precisely, we obtain the correct order of magnitude upper and lower bounds for the set of distinct even legs with 5-almost prime hypotenuse. This is a strong version of the appropriate analogy to a conjecture of Chow and Pomerance (stated there for prime hypotenuse).

2401.16769

Ming Ji

Ming Ji, Jonte R. Hance, and Holger F. Hofmann

Tracing quantum correlations back to collective interferences

10 pages, 4 figures

10.1088/1367-2630/ad5619

quant-ph physics.hist-ph

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

In this paper, we investigate the possibility of explaining nonclassical correlations between two quantum systems in terms of quantum interferences between collective states of the two systems. We achieve this by mapping the relations between different measurement contexts in the product Hilbert space of a pair of two-level systems onto an analogous sequence of interferences between paths in a single-particle interferometer. The relations between different measurement outcomes are then traced to the distribution of probability currents in the interferometer, where paradoxical relations between the outcomes are identified with currents connecting two states that are orthogonal and should therefore exclude each other. We show that the relation between probability currents and correlations can be represented by continuous conditional (quasi)probability currents through the interferometer, given by weak values; the violation of the noncontextual assumption is expressed by negative conditional currents in some of the paths. Since negative conditional currents correspond to the assignment of negative conditional probabilities to measurements results in different measurement contexts, the necessity of such negative probability currents represents a failure of noncontextual local realism. Our results help to explain the meaning of nonlocal correlations in quantum mechanics, and support Feynman's claim that interference is the origin of all quantum phenomena.

2401.16812

Takaaki Saeki

Takaaki Saeki, Soumi Maiti, Shinnosuke Takamichi, Shinji Watanabe, Hiroshi Saruwatari

SpeechBERTScore: Reference-Aware Automatic Evaluation of Speech Generation Leveraging NLP Evaluation Metrics

Accepted by Interspeech 2024. An extended version with Appendix. Code: https://github.com/Takaaki-Saeki/DiscreteSpeechMetrics

cs.SD eess.AS

http://creativecommons.org/licenses/by-sa/4.0/

While subjective assessments have been the gold standard for evaluating speech generation, there is a growing need for objective metrics that are highly correlated with human subjective judgments due to their cost efficiency. This paper proposes reference-aware automatic evaluation methods for speech generation inspired by evaluation metrics in natural language processing. The proposed SpeechBERTScore computes the BERTScore for self-supervised dense speech features of the generated and reference speech, which can have different sequential lengths. We also propose SpeechBLEU and SpeechTokenDistance, which are computed on speech discrete tokens. The evaluations on synthesized speech show that our method correlates better with human subjective ratings than mel cepstral distortion and a recent mean opinion score prediction model. Also, they are effective in noisy speech evaluation and have cross-lingual applicability.

2401.17230

Jee-Weon Jung

Jee-weon Jung, Wangyou Zhang, Jiatong Shi, Zakaria Aldeneh, Takuya Higuchi, Barry-John Theobald, Ahmed Hussen Abdelaziz, Shinji Watanabe

ESPnet-SPK: full pipeline speaker embedding toolkit with reproducible recipes, self-supervised front-ends, and off-the-shelf models

5 pages, 3 figures, 7 tables, Interspeech 2024

cs.SD cs.AI eess.AS

http://creativecommons.org/licenses/by/4.0/

This paper introduces ESPnet-SPK, a toolkit designed with several objectives for training speaker embedding extractors. First, we provide an open-source platform for researchers in the speaker recognition community to effortlessly build models. We provide several models, ranging from x-vector to recent SKA-TDNN. Through the modularized architecture design, variants can be developed easily. We also aspire to bridge developed models with other domains, facilitating the broad research community to effortlessly incorporate state-of-the-art embedding extractors. Pre-trained embedding extractors can be accessed in an off-the-shelf manner and we demonstrate the toolkit's versatility by showcasing its integration with two tasks. Another goal is to integrate with diverse self-supervised learning features. We release a reproducible recipe that achieves an equal error rate of 0.39% on the Vox1-O evaluation protocol using WavLM-Large with ECAPA-TDNN.

2401.17619

Jiatong Shi

Jiatong Shi, Yueqian Lin, Xinyi Bai, Keyi Zhang, Yuning Wu, Yuxun Tang, Yifeng Yu, Qin Jin, Shinji Watanabe

Singing Voice Data Scaling-up: An Introduction to ACE-Opencpop and ACE-KiSing

Accepted by Interspeech2024

cs.SD eess.AS

http://creativecommons.org/licenses/by-sa/4.0/

In singing voice synthesis (SVS), generating singing voices from musical scores faces challenges due to limited data availability. This study proposes a unique strategy to address the data scarcity in SVS. We employ an existing singing voice synthesizer for data augmentation, complemented by detailed manual tuning, an approach not previously explored in data curation, to reduce instances of unnatural voice synthesis. This innovative method has led to the creation of two expansive singing voice datasets, ACE-Opencpop and ACE-KiSing, which are instrumental for large-scale, multi-singer voice synthesis. Through thorough experimentation, we establish that these datasets not only serve as new benchmarks for SVS but also enhance SVS performance on other singing voice datasets when used as supplementary resources. The corpora, pre-trained models, and their related training recipes are publicly available at ESPnet-Muskits (\url{https://github.com/espnet/espnet})

2401.17731

Amon Lahr

Amon Lahr, Filip Tronarp, Nathanael Bosch, Jonathan Schmidt, Philipp Hennig, Melanie N. Zeilinger

Probabilistic ODE Solvers for Integration Error-Aware Numerical Optimal Control

to be published in the 6th Annual Learning for Dynamics & Control Conference (L4DC 2024)

math.OC

http://creativecommons.org/licenses/by/4.0/

Appropriate time discretization is crucial for real-time applications of numerical optimal control, such as nonlinear model predictive control. However, if the discretization error strongly depends on the applied control input, meeting accuracy and sampling time requirements simultaneously can be challenging using classical discretization methods. In particular, neither fixed-grid nor adaptive-grid discretizations may be suitable, when they suffer from large integration error or exceed the prescribed sampling time, respectively. In this work, we take a first step at closing this gap by utilizing probabilistic numerical integrators to approximate the solution of the initial value problem, as well as the computational uncertainty associated with it, inside the optimal control problem (OCP). By taking the viewpoint of probabilistic numerics and propagating the numerical uncertainty in the cost, the OCP is reformulated such that the optimal input reduces the computational uncertainty insofar as it is beneficial for the control objective. The proposed approach is illustrated using a numerical example, and potential benefits and limitations are discussed.

2402.00340

Zakaria Aldeneh

Zakaria Aldeneh, Takuya Higuchi, Jee-weon Jung, Skyler Seto, Tatiana Likhomanenko, Stephen Shum, Ahmed Hussen Abdelaziz, Shinji Watanabe, Barry-John Theobald

Can you Remove the Downstream Model for Speaker Recognition with Self-Supervised Speech Features?

cs.SD eess.AS

http://creativecommons.org/licenses/by/4.0/

Self-supervised features are typically used in place of filter-bank features in speaker verification models. However, these models were originally designed to ingest filter-bank features as inputs, and thus, training them on top of self-supervised features assumes that both feature types require the same amount of learning for the task. In this work, we observe that pre-trained self-supervised speech features inherently include information required for downstream speaker verification task, and therefore, we can simplify the downstream model without sacrificing performance. To this end, we revisit the design of the downstream model for speaker verification using self-supervised features. We show that we can simplify the model to use 97.51% fewer parameters while achieving a 29.93% average improvement in performance on SUPERB. Consequently, we show that the simplified downstream model is more data efficient compared to baseline--it achieves better performance with only 60% of the training data.

2402.00453

Tosin Adewumi

Tosin Adewumi, Nudrat Habib, Lama Alkhaled, and Elisa Barney

Instruction Makes a Difference

Accepted at the 16th IAPR International Workshop On Document Analysis Systems (DAS)

cs.CV cs.CL

http://creativecommons.org/licenses/by/4.0/

We introduce Instruction Document Visual Question Answering (iDocVQA) dataset and Large Language Document (LLaDoc) model, for training Language-Vision (LV) models for document analysis and predictions on document images, respectively. Usually, deep neural networks for the DocVQA task are trained on datasets lacking instructions. We show that using instruction-following datasets improves performance. We compare performance across document-related datasets using the recent state-of-the-art (SotA) Large Language and Vision Assistant (LLaVA)1.5 as the base model. We also evaluate the performance of the derived models for object hallucination using the Polling-based Object Probing Evaluation (POPE) dataset. The results show that instruction-tuning performance ranges from 11X to 32X of zero-shot performance and from 0.1% to 4.2% over non-instruction (traditional task) finetuning. Despite the gains, these still fall short of human performance (94.36%), implying there's much room for improvement.

2402.00604

Andrew Robert Clifford Spiers A R C S

Andrew Spiers

Analytically Separating the Source of the Teukolsky Equation

17 pages, 5 figures, accompanying Mathematica notebook available at https://github.com/DrAndrewSpiers/Teukolsky-equation-source-decomposer Formerly titled: "Efficiently Separating the Source of the Teukolsky Equation"

gr-qc

http://creativecommons.org/licenses/by-sa/4.0/

Recent gravitational wave detections from black hole mergers have underscored the critical role black hole perturbation theory and the Teukolsky equation play in understanding the behaviour of black holes. The separable nature of the Teukolsky equation has long been leveraged to study the vacuum linear Teukolsky equation; however, as theory and measurements advance, solving the sourced Teukolsky equation is becoming a frontier of research. In particular, second-order calculations, such as in quasi-normal mode and self-force problems, have extended sources. This paper presents a novel method for analytically separating the Teukolsky equation's source, aimed to improve efficiency. Separating the source is a non-trivial problem due to the angular and radial mixing of generic quantities in Kerr spacetime. We provide a proof-of-concept demonstration of our method and show that it is accurate, separating the Teukolsky source produced by the stress-energy tensor of an ideal gas cloud surrounding a Kerr black hole. The detailed application of our method is provided in an accompanying \textit{Mathematica} notebook. Our approach opens up a new avenue for accurate black hole perturbation theory calculations with sources in both the time and frequency domain.

2402.00811

Bunlong Lay

Bunlong Lay, Timo Gerkmann

An Analysis of the Variance of Diffusion-based Speech Enhancement

5 pages, 3 figures, 1 table

eess.AS cs.LG cs.SD

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Diffusion models proved to be powerful models for generative speech enhancement. In recent SGMSE+ approaches, training involves a stochastic differential equation for the diffusion process, adding both Gaussian and environmental noise to the clean speech signal gradually. The speech enhancement performance varies depending on the choice of the stochastic differential equation that controls the evolution of the mean and the variance along the diffusion processes when adding environmental and Gaussian noise. In this work, we highlight that the scale of the variance is a dominant parameter for speech enhancement performance and show that it controls the tradeoff between noise attenuation and speech distortions. More concretely, we show that a larger variance increases the noise attenuation and allows for reducing the computational footprint, as fewer function evaluations for generating the estimate are required

2402.01460

Ruoxuan Li

Jinyuan Chang, Zhao Ding, Yuling Jiao, Ruoxuan Li, Jerry Zhijian Yang

Deep conditional distribution learning via conditional F\"ollmer flow

The original title of this paper is "Deep Conditional Generative Learning: Model and Error Analysis"

stat.ML cs.LG

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We introduce an ordinary differential equation (ODE) based deep generative method for learning conditional distributions, named Conditional F\"ollmer Flow. Starting from a standard Gaussian distribution, the proposed flow could approximate the target conditional distribution very well when the time is close to 1. For effective implementation, we discretize the flow with Euler's method where we estimate the velocity field nonparametrically using a deep neural network. Furthermore, we also establish the convergence result for the Wasserstein-2 distance between the distribution of the learned samples and the target conditional distribution, providing the first comprehensive end-to-end error analysis for conditional distribution learning via ODE flow. Our numerical experiments showcase its effectiveness across a range of scenarios, from standard nonparametric conditional density estimation problems to more intricate challenges involving image data, illustrating its superiority over various existing conditional density estimation methods.

2402.02100

Rong-Chun Ge

Ling Ye, Lan Luo, An Wang, Rongchun Ge, Zhiyou Zhang

Weak-measurement-based pseudospin pointer: A cost-effective scheme for precision measurement

6 figures

Phys. Rev. A 109, L060601 (2024)

10.1103/PhysRevA.109.L060601

quant-ph physics.optics

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

As an essential component of state-of-the-art quantum technologies, fast and efficient quantum measurements are in persistent demand over time. We present a proof-of-principle experiment on a new dimensionless pseudo-spin pointer based on weak measurement. In the context of optical parameter estimation, we demonstrate that the parametric distribution's moment is obtained experimentally by employing the dimensionless pointer without measuring the distribution literally. In addition to the sheer liberation of experimental expense, the photon-countering-based pointer is well-calibrated for the detection of weak signals. We show that for signals $3$-$4$ orders of weaker in strength than the area-array camera method, an order of improvement in precision is achieved experimentally.

2402.02520

Linn\'ea Gyllingberg

Linn\'ea Gyllingberg, Yu Tian, and David J.T. Sumpter

A minimal model of cognition based on oscillatory and reinforcement processes

q-bio.NC cs.SI math.DS nlin.AO physics.bio-ph

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Building mathematical models of brains is difficult because of the sheer complexity of the problem. One potential starting point is through basal cognition, which give abstract representation of a range of organisms without central nervous systems, including fungi, slime moulds and bacteria. We propose one such model, demonstrating how a combination of oscillatory and current-based reinforcement processes can be used to couple resources in an efficient manner, mimicking the way these organisms function. A key ingredient in our model, not found in previous basal cognition models, is that we explicitly model oscillations in the number of particles (i.e. the nutrients, chemical signals or similar, which make up the biological system) and the flow of these particles within the modelled organisms. Using this approach, we find that our model builds efficient solutions, provided the environmental oscillations are sufficiently out of phase. We further demonstrate that amplitude differences can promote efficient solutions and that the system is robust to frequency differences. In the context of these findings, we discuss connections between our model and basal cognition in biological systems and slime moulds, in particular, how oscillations might contribute to self-organised problem-solving by these organisms.

2402.03293

Yongchang Hao

Yongchang Hao, Yanshuai Cao, Lili Mou

Flora: Low-Rank Adapters Are Secretly Gradient Compressors

Accepted @ ICML 2024

cs.LG cs.AI stat.ML

http://creativecommons.org/licenses/by-nc-sa/4.0/

Despite large neural networks demonstrating remarkable abilities to complete different tasks, they require excessive memory usage to store the optimization states for training. To alleviate this, the low-rank adaptation (LoRA) is proposed to reduce the optimization states by training fewer parameters. However, LoRA restricts overall weight update matrices to be low-rank, limiting the model performance. In this work, we investigate the dynamics of LoRA and identify that it can be approximated by a random projection. Based on this observation, we propose Flora, which is able to achieve high-rank updates by resampling the projection matrices while enjoying the sublinear space complexity of optimization states. We conduct experiments across different tasks and model architectures to verify the effectiveness of our approach.

2402.03608

Selim M. Shahriar

Jinyang Li, Timothy Kovachy, Jason Bonacum, Selim M. Shahriar

Sensitivity and Bandwidth of a Point-Source-Interferometry-based Inertial Measurement Unit Employing Large Momentum Transfer and Launched Atoms

Atoms 2024, 12(6), 32

10.3390/atoms12060032

quant-ph

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We analyze theoretically the sensitivity and bandwidth of accelerometry and rotation sensing with a point source interferometer employing large momentum transfer (LMT) with molasses-launched atoms. The launching process makes it possible to realize the LMT process without the need to physically change directions of the Raman pulses, thus significantly simplifying the apparatus and reducing the amount of time needed to make the measurements. These advantages become more important when this process is used for realizing an inertial measurement unit (IMU) that can measure rotation around and acceleration along each of the three axes. We describe an explicit scheme for a such an IMU and determine the expected sensitivity and bandwidth thereof for experimentally accessible parameters.

2402.03915

Olivier Jeunen

Olivier Jeunen and Aleksei Ustimenko

Learning Metrics that Maximise Power for Accelerated A/B-Tests

To appear in the Applied Data Science track at the ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD '24)

cs.LG cs.IR stat.AP stat.ML

http://creativecommons.org/licenses/by/4.0/

Online controlled experiments are a crucial tool to allow for confident decision-making in technology companies. A North Star metric is defined (such as long-term revenue or user retention), and system variants that statistically significantly improve on this metric in an A/B-test can be considered superior. North Star metrics are typically delayed and insensitive. As a result, the cost of experimentation is high: experiments need to run for a long time, and even then, type-II errors (i.e. false negatives) are prevalent. We propose to tackle this by learning metrics from short-term signals that directly maximise the statistical power they harness with respect to the North Star. We show that existing approaches are prone to overfitting, in that higher average metric sensitivity does not imply improved type-II errors, and propose to instead minimise the $p$-values a metric would have produced on a log of past experiments. We collect such datasets from two social media applications with over 160 million Monthly Active Users each, totalling over 153 A/B-pairs. Empirical results show that we are able to increase statistical power by up to 78% when using our learnt metrics stand-alone, and by up to 210% when used in tandem with the North Star. Alternatively, we can obtain constant statistical power at a sample size that is down to 12% of what the North Star requires, significantly reducing the cost of experimentation.

2402.04276

Anh-Vu Phan

J. D. Phan and A.-V. Phan

Accelerated boundary integral analysis of energy eigenvalues for confined electron states in quantum semiconductor heterostructures

cond-mat.mes-hall math-ph math.MP

http://creativecommons.org/licenses/by-nc-nd/4.0/

This paper presents a novel and efficient approach for the computation of energy eigenvalues in quantum semiconductor heterostructures. Accurate determination of the electronic states in these heterostructures is crucial for understanding their optical and electronic properties, making it a key challenge in semiconductor physics. The proposed method is based on utilizing series expansions of zero-order Bessel functions to numerically solve the Schr\"odinger equation using boundary integral method for bound electron states in a computationally efficient manner. To validate the proposed technique, we applied it to address previously explored issues by other research groups. The results clearly demonstrate the computational efficiency and high precision of our approach. Notably, the proposed technique significantly reduces the computational time compared to the conventional method for searching the energy eigenvalues in quantum structures.

2402.04278

Seongsu Kim

Seongsu Kim, Sungsoo Ahn

Gaussian Plane-Wave Neural Operator for Electron Density Estimation

Accepted by ICML 2024 main poster

International Conference on Machine Learning (ICML), 2024

physics.chem-ph cs.LG

http://creativecommons.org/licenses/by/4.0/

This work studies machine learning for electron density prediction, which is fundamental for understanding chemical systems and density functional theory (DFT) simulations. To this end, we introduce the Gaussian plane-wave neural operator (GPWNO), which operates in the infinite-dimensional functional space using the plane-wave and Gaussian-type orbital bases, widely recognized in the context of DFT. In particular, both high- and low-frequency components of the density can be effectively represented due to the complementary nature of the two bases. Extensive experiments on QM9, MD, and material project datasets demonstrate GPWNO's superior performance over ten baselines.

2402.04333

Mengzhou Xia

Mengzhou Xia, Sadhika Malladi, Suchin Gururangan, Sanjeev Arora, Danqi Chen

LESS: Selecting Influential Data for Targeted Instruction Tuning

ICML 2024; Code and data are available at https://github.com/princeton-nlp/LESS

cs.CL cs.AI cs.LG

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Instruction tuning has unlocked powerful capabilities in large language models (LLMs), effectively using combined datasets to develop generalpurpose chatbots. However, real-world applications often require a specialized suite of skills (e.g., reasoning). The challenge lies in identifying the most relevant data from these extensive datasets to effectively develop specific capabilities, a setting we frame as targeted instruction tuning. We propose LESS, an optimizer-aware and practically efficient algorithm to effectively estimate data influences and perform Low-rank gradiEnt Similarity Search for instruction data selection. Crucially, LESS adapts existing influence formulations to work with the Adam optimizer and variable-length instruction data. LESS first constructs a highly reusable and transferable gradient datastore with low-dimensional gradient features and then selects examples based on their similarity to few-shot examples embodying a specific capability. Experiments show that training on a LESS-selected 5% of the data can often outperform training on the full dataset across diverse downstream tasks. Furthermore, the selected data is highly transferable: smaller models can be leveraged to select useful data for larger models and models from different families. Our qualitative analysis shows that our method goes beyond surface form cues to identify data that exemplifies the necessary reasoning skills for the intended downstream application.

2402.04655

Shuoyuan Wang

Shuoyuan Wang, Jindong Wang, Guoqing Wang, Bob Zhang, Kaiyang Zhou, Hongxin Wei

Open-Vocabulary Calibration for Fine-tuned CLIP

Accepted by ICML 2024

cs.LG

http://creativecommons.org/publicdomain/zero/1.0/

Vision-language models (VLMs) have emerged as formidable tools, showing their strong capability in handling various open-vocabulary tasks in image recognition, text-driven visual content generation, and visual chatbots, to name a few. In recent years, considerable efforts and resources have been devoted to adaptation methods for improving downstream performance of VLMs, particularly on parameter-efficient fine-tuning methods like prompt learning. However, a crucial aspect that has been largely overlooked is the confidence calibration problem in fine-tuned VLMs, which could greatly reduce reliability when deploying such models in the real world. This paper bridges the gap by systematically investigating the confidence calibration problem in the context of prompt learning and reveals that existing calibration methods are insufficient to address the problem, especially in the open-vocabulary setting. To solve the problem, we present a simple and effective approach called Distance-Aware Calibration (DAC), which is based on scaling the temperature using as guidance the distance between predicted text labels and base classes. The experiments with 7 distinct prompt learning methods applied across 11 diverse downstream datasets demonstrate the effectiveness of DAC, which achieves high efficacy without sacrificing the inference speed. Our code is available at [this https URL](https://github.com/ml-stat-Sustech/CLIP_Calibration).

2402.05482

Cho-Yuan Lee

Cho-Yuan Lee, Kuan-Chen Wang, Kai-Chun Liu, Yu-Te Wang, Xugang Lu, Ping-Cheng Yeh, and Yu Tsao

A Non-Intrusive Neural Quality Assessment Model for Surface Electromyography Signals

5 pages, 4 figures

eess.SP cs.LG

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

In practical scenarios involving the measurement of surface electromyography (sEMG) in muscles, particularly those areas near the heart, one of the primary sources of contamination is the presence of electrocardiogram (ECG) signals. To assess the quality of real-world sEMG data more effectively, this study proposes QASE-net, a new non-intrusive model that predicts the SNR of sEMG signals. QASE-net combines CNN-BLSTM with attention mechanisms and follows an end-to-end training strategy. Our experimental framework utilizes real-world sEMG and ECG data from two open-access databases, the Non-Invasive Adaptive Prosthetics Database and the MIT-BIH Normal Sinus Rhythm Database, respectively. The experimental results demonstrate the superiority of QASE-net over the previous assessment model, exhibiting significantly reduced prediction errors and notably higher linear correlations with the ground truth. These findings show the potential of QASE-net to substantially enhance the reliability and precision of sEMG quality assessment in practical applications.

2402.05689

Yige Hong

Yige Hong, Qiaomin Xie, Yudong Chen, Weina Wang

Unichain and Aperiodicity are Sufficient for Asymptotic Optimality of Average-Reward Restless Bandits

49 pages, 3 figures. This version adds details on the unichain condition, stationary distribution, and long-run time average

cs.LG math.OC math.PR

http://creativecommons.org/licenses/by/4.0/

We consider the infinite-horizon, average-reward restless bandit problem in discrete time. We propose a new class of policies that are designed to drive a progressively larger subset of arms toward the optimal distribution. We show that our policies are asymptotically optimal with an $O(1/\sqrt{N})$ optimality gap for an $N$-armed problem, provided that the single-armed MDP is unichain and aperiodic under the optimal single-armed policy. Our approach departs from most existing work that focuses on index or priority policies, which rely on the Uniform Global Attractor Property (UGAP) to guarantee convergence to the optimum, or a recently developed simulation-based policy, which requires a Synchronization Assumption (SA).

2402.05889

Shoubin Yu

Shoubin Yu, Jaehong Yoon, Mohit Bansal

CREMA: Generalizable and Efficient Video-Language Reasoning via Multimodal Modular Fusion

first two authors contributed equally. Project page: https://CREMA-VideoLLM.github.io/

cs.CV cs.AI cs.CL

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Despite impressive advancements in recent multimodal reasoning approaches, they are still limited in flexibility and efficiency, as these models typically process only a few fixed modality inputs and require updates to numerous parameters. This paper tackles these critical challenges and proposes CREMA, a generalizable, highly efficient, and modular modality-fusion framework that can incorporate any new modality to enhance video reasoning. We first augment multiple informative modalities (such as optical flow, 3D point cloud, audio, thermal heatmap, and touch map) from given videos without extra human annotation by leveraging sensors or existing pre-trained models. Next, we introduce a query transformer with multiple parameter-efficient modules associated with each accessible modality. It projects diverse modality features to the LLM token embedding space, allowing the model to integrate different data types for response generation. Furthermore, we propose a novel progressive multimodal fusion design supported by a lightweight fusion module and modality-sequential training strategy. It helps compress information across various assisting modalities, maintaining computational efficiency in the LLM while improving performance. We validate our method on 7 video-language reasoning tasks assisted by diverse modalities, including conventional VideoQA and Video-Audio/3D/Touch/Thermal QA, and achieve better/equivalent performance against strong multimodal LLMs, including OneLLM, BLIP-2, and SeViLA while reducing over 90% trainable parameters. We provide extensive analyses of CREMA, including the impact of each modality on reasoning domains, the design of the fusion module, and example visualizations.

2402.05928

Ingvar Ziemann

Ingvar Ziemann, Stephen Tu, George J. Pappas, Nikolai Matni

Sharp Rates in Dependent Learning Theory: Avoiding Sample Size Deflation for the Square Loss

cs.LG stat.ML

http://creativecommons.org/licenses/by/4.0/

In this work, we study statistical learning with dependent ($\beta$-mixing) data and square loss in a hypothesis class $\mathscr{F}\subset L_{\Psi_p}$ where $\Psi_p$ is the norm $\|f\|_{\Psi_p} \triangleq \sup_{m\geq 1} m^{-1/p} \|f\|_{L^m} $ for some $p\in [2,\infty]$. Our inquiry is motivated by the search for a sharp noise interaction term, or variance proxy, in learning with dependent data. Absent any realizability assumption, typical non-asymptotic results exhibit variance proxies that are deflated multiplicatively by the mixing time of the underlying covariates process. We show that whenever the topologies of $L^2$ and $\Psi_p$ are comparable on our hypothesis class $\mathscr{F}$ -- that is, $\mathscr{F}$ is a weakly sub-Gaussian class: $\|f\|_{\Psi_p} \lesssim \|f\|_{L^2}^\eta$ for some $\eta\in (0,1]$ -- the empirical risk minimizer achieves a rate that only depends on the complexity of the class and second order statistics in its leading term. Our result holds whether the problem is realizable or not and we refer to this as a \emph{near mixing-free rate}, since direct dependence on mixing is relegated to an additive higher order term. We arrive at our result by combining the above notion of a weakly sub-Gaussian class with mixed tail generic chaining. This combination allows us to compute sharp, instance-optimal rates for a wide range of problems. Examples that satisfy our framework include sub-Gaussian linear regression, more general smoothly parameterized function classes, finite hypothesis classes, and bounded smoothness classes.

2402.06160

Jongha Jon Ryu

Maohao Shen, J. Jon Ryu, Soumya Ghosh, Yuheng Bu, Prasanna Sattigeri, Subhro Das, Gregory W. Wornell

Are Uncertainty Quantification Capabilities of Evidential Deep Learning a Mirage?

29 pages, 12 figures

cs.LG stat.ML

http://creativecommons.org/licenses/by/4.0/

This paper questions the effectiveness of a modern predictive uncertainty quantification approach, called \emph{evidential deep learning} (EDL), in which a single neural network model is trained to learn a meta distribution over the predictive distribution by minimizing a specific objective function. Despite their perceived strong empirical performance on downstream tasks, a line of recent studies by Bengs et al. identify limitations of the existing methods to conclude their learned epistemic uncertainties are unreliable, e.g., in that they are non-vanishing even with infinite data. Building on and sharpening such analysis, we 1) provide a sharper understanding of the asymptotic behavior of a wide class of EDL methods by unifying various objective functions; 2) reveal that the EDL methods can be better interpreted as an out-of-distribution detection algorithm based on energy-based-models; and 3) conduct extensive ablation studies to better assess their empirical effectiveness with real-world datasets. Through all these analyses, we conclude that even when EDL methods are empirically effective on downstream tasks, this occurs despite their poor uncertainty quantification capabilities. Our investigation suggests that incorporating model uncertainty can help EDL methods faithfully quantify uncertainties and further improve performance on representative downstream tasks, albeit at the cost of additional computational complexity.

2402.06339

Sam Nerenberg

Sam Nerenberg, Oliver D. Neill, Giulia Marcucci and Daniele Faccio

Photon Number-Resolving Quantum Reservoir Computing

Information from supplementary migrated to main body. Some rewording and new figures. Reformatted

quant-ph physics.app-ph physics.optics

http://creativecommons.org/licenses/by/4.0/

Neuromorphic processors improve the efficiency of machine learning algorithms through the implementation of physical artificial neurons to perform computations. However, whilst efficient classical neuromorphic processors have been demonstrated in various forms, practical quantum neuromorphic platforms are still in the early stages of development. Here we propose a fixed optical network for photonic quantum reservoir computing that is enabled by photon number-resolved detection of the output states. This significantly reduces the required complexity of the input quantum states while still accessing a high-dimensional Hilbert space. The approach is implementable with currently available technology and lowers the barrier to entry to quantum machine learning.

2402.06784

Stefano Martina PhD

Matteo Paiano, Stefano Martina, Carlotta Giannelli, Filippo Caruso

Transfer learning with generative models for object detection on limited datasets

28 pages, 16 figures, 1 table

cs.CV cond-mat.dis-nn cs.AI cs.LG cs.NA math.NA

http://creativecommons.org/licenses/by/4.0/

The availability of data is limited in some fields, especially for object detection tasks, where it is necessary to have correctly labeled bounding boxes around each object. A notable example of such data scarcity is found in the domain of marine biology, where it is useful to develop methods to automatically detect submarine species for environmental monitoring. To address this data limitation, the state-of-the-art machine learning strategies employ two main approaches. The first involves pretraining models on existing datasets before generalizing to the specific domain of interest. The second strategy is to create synthetic datasets specifically tailored to the target domain using methods like copy-paste techniques or ad-hoc simulators. The first strategy often faces a significant domain shift, while the second demands custom solutions crafted for the specific task. In response to these challenges, here we propose a transfer learning framework that is valid for a generic scenario. In this framework, generated images help to improve the performances of an object detector in a few-real data regime. This is achieved through a diffusion-based generative model that was pretrained on large generic datasets. With respect to the state-of-the-art, we find that it is not necessary to fine tune the generative model on the specific domain of interest. We believe that this is an important advance because it mitigates the labor-intensive task of manual labeling the images in object detection tasks. We validate our approach focusing on fishes in an underwater environment, and on the more common domain of cars in an urban setting. Our method achieves detection performance comparable to models trained on thousands of images, using only a few hundreds of input data. Our results pave the way for new generative AI-based protocols for machine learning applications in various domains.

2402.07061

Hung P. Hoang

Sophia Heimann, Hung P. Hoang, Stefan Hougardy

The $k$-Opt algorithm for the Traveling Salesman Problem has exponential running time for $k \ge 5$

Appeared in ICALP 2024

cs.DS cs.DM math.CO

http://creativecommons.org/licenses/by-nc-nd/4.0/

The $k$-Opt algorithm is a local search algorithm for the Traveling Salesman Problem. Starting with an initial tour, it iteratively replaces at most $k$ edges in the tour with the same number of edges to obtain a better tour. Krentel (FOCS 1989) showed that the Traveling Salesman Problem with the $k$-Opt neighborhood is complete for the class PLS (polynomial time local search) and that the $k$-Opt algorithm can have exponential running time for any pivot rule. However, his proof requires $k \gg 1000$ and has a substantial gap. We show the two properties above for a much smaller value of $k$, addressing an open question by Monien, Dumrauf, and Tscheuschner (ICALP 2010). In particular, we prove the PLS-completeness for $k \geq 17$ and the exponential running time for $k \geq 5$.

2402.07586

Teresa Salazar

Teresa Salazar and Jo\~ao Gama and Helder Ara\'ujo and Pedro Henriques Abreu

Unveiling Group-Specific Distributed Concept Drift: A Fairness Imperative in Federated Learning

cs.LG

http://creativecommons.org/licenses/by/4.0/

In the evolving field of machine learning, ensuring fairness has become a critical concern, prompting the development of algorithms designed to mitigate discriminatory outcomes in decision-making processes. However, achieving fairness in the presence of group-specific concept drift remains an unexplored frontier, and our research represents pioneering efforts in this regard. Group-specific concept drift refers to situations where one group experiences concept drift over time while another does not, leading to a decrease in fairness even if accuracy remains fairly stable. Within the framework of federated learning, where clients collaboratively train models, its distributed nature further amplifies these challenges since each client can experience group-specific concept drift independently while still sharing the same underlying concept, creating a complex and dynamic environment for maintaining fairness. One of the significant contributions of our research is the formalization and introduction of the problem of group-specific concept drift and its distributed counterpart, shedding light on its critical importance in the realm of fairness. In addition, leveraging insights from prior research, we adapt an existing distributed concept drift adaptation algorithm to tackle group-specific distributed concept drift which utilizes a multi-model approach, a local group-specific drift detection mechanism, and continuous clustering of models over time. The findings from our experiments highlight the importance of addressing group-specific concept drift and its distributed counterpart to advance fairness in machine learning.

2402.08668

The CMS Collaboration

CMS Collaboration

Search for pair production of scalar and vector leptoquarks decaying to muons and bottom quarks in proton-proton collisions at $\sqrt{s}$ = 13 TeV

Replaced with the published version. Added the journal reference and the DOI. All the figures and tables can be found at http://cms-results.web.cern.ch/cms-results/public-results/publications/EXO-21-019 (CMS Public Pages)

Phys. Rev. D 109 (2024) 112003

10.1103/PhysRevD.109.112003

CMS-EXO-21-019, CERN-EP-2023-301

hep-ex

http://creativecommons.org/licenses/by/4.0/

A search for pair production of scalar and vector leptoquarks (LQs) each decaying to a muon and a bottom quark is performed using proton-proton collision data collected at $\sqrt{s}$ = 13 TeV with the CMS detector at the CERN LHC, corresponding to an integrated luminosity of 138 fb$^{-1}$. No excess above standard model expectation is observed. Scalar (vector) LQs with masses less than 1810 (2120) GeV are excluded at 95% confidence level, assuming a 100% branching fraction of the LQ decaying to a muon and a bottom quark. These limits represent the most stringent to date.

2402.09337

Luis Altenkort

Luis Altenkort, David de la Cruz, Olaf Kaczmarek, Guy D. Moore, Hai-Tao Shu

Lattice B-field correlators for heavy quarks

12 pages, 9 figures

Phys.Rev.D 109 (2024) 11, 114505

10.1103/PhysRevD.109.114505

hep-lat hep-ph

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We analyze the color-magnetic (or "$B$") field two-point function that encodes the finite-mass correction to the heavy quark momentum diffusion coefficient. The simulations are done on fine isotropic lattices in the quenched approximation at $1.5\,T_c$, using a range of gradient flow times for noise suppression and operator renormalization. The continuum extrapolation is performed at fixed flow time followed by a second extrapolation to zero flow time. Perturbative calculations to next-to-leading order of this correlation function, matching gradient-flowed correlators to MS-bar, are used to resolve nontrivial renormalization issues. We perform a spectral reconstruction based on perturbative model fits to estimate the coefficient $\kappa_B$ of the finite-mass correction to the heavy quark momentum diffusion coefficient. The approach we present here yields high-precision data for the correlator with all renormalization issues incorporated at next-to-leading order, and is also applicable for actions with dynamical fermions.

2402.09857

Laura Hunt

Laura. J. Hunt, Kevin. A. Pimbblet, and David. M. Benoit

Predicting the Ages of Galaxies with an Artificial Neural Network

20 pages, 19 figures, Published in MNRAS - please use arXiv version

astro-ph.GA

http://creativecommons.org/licenses/by/4.0/

We present a new method of predicting the ages of galaxies using a machine learning (ML) algorithm with the goal of providing an alternative to traditional methods. We aim to match the ability of traditional models to predict the ages of galaxies by training an artificial neural network (ANN) to recognise the relationships between the equivalent widths of spectral indices and the mass-weighted ages of galaxies estimated by the MAGPHYS model in data release 3 (DR3) of the Galaxy and Mass Assembly (GAMA) survey. We discuss the optimisation of our hyperparameters extensively and investigate the application of a custom loss function to reduce the influence of errors in our input data. To quantify the quality of our predictions we calculate the mean squared error (MSE), mean absolute error (MAE) and R^2 score for which we find MSE = 0.020, MAE = 0.108 and R^2 = 0.530. We find our predicted ages have a similar distribution with standard deviation sigma_p = 0.182 compared with the GAMA true ages sigma_t = 0.207. This is achieved in approximately 23s to train our ANN on an 11th Gen Intel Core i9-11900H running at 2.50GHz using 32GB of RAM. We report our results for when light-weighted ages are used to train the ANN, which improves the accuracy of the predictions. Finally, we detail an evaluation of our method relating to physical properties and compare with other ML techniques to encourage future applications of ML techniques in Astronomy.

2402.09871

Zihao Wang

Zihao Wang, Shuyu Li, Tao Zhang, Qi Wang, Pengfei Yu, Jinyang Luo, Yan Liu, Ming Xi, Kejun Zhang

MuChin: A Chinese Colloquial Description Benchmark for Evaluating Language Models in the Field of Music

Accepted by International Joint Conference on Artificial Intelligence 2024 (IJCAI 2024)

cs.SD cs.AI cs.MM eess.AS

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

The rapidly evolving multimodal Large Language Models (LLMs) urgently require new benchmarks to uniformly evaluate their performance on understanding and textually describing music. However, due to semantic gaps between Music Information Retrieval (MIR) algorithms and human understanding, discrepancies between professionals and the public, and low precision of annotations, existing music description datasets cannot serve as benchmarks. To this end, we present MuChin, the first open-source music description benchmark in Chinese colloquial language, designed to evaluate the performance of multimodal LLMs in understanding and describing music. We established the Caichong Music Annotation Platform (CaiMAP) that employs an innovative multi-person, multi-stage assurance method, and recruited both amateurs and professionals to ensure the precision of annotations and alignment with popular semantics. Utilizing this method, we built a dataset with multi-dimensional, high-precision music annotations, the Caichong Music Dataset (CaiMD), and carefully selected 1,000 high-quality entries to serve as the test set for MuChin. Based on MuChin, we analyzed the discrepancies between professionals and amateurs in terms of music description, and empirically demonstrated the effectiveness of annotated data for fine-tuning LLMs. Ultimately, we employed MuChin to evaluate existing music understanding models on their ability to provide colloquial descriptions of music. All data related to the benchmark, along with the scoring code and detailed appendices, have been open-sourced (https://github.com/CarlWangChina/MuChin/).

2402.10036

Anastasios Tsiamis

Anastasios Tsiamis, Aren Karapetyan, Yueshan Li, Efe C. Balta, John Lygeros

Predictive Linear Online Tracking for Unknown Targets

ICML 2024 (spotlight)

eess.SY cs.LG cs.SY math.OC

http://creativecommons.org/licenses/by/4.0/

In this paper, we study the problem of online tracking in linear control systems, where the objective is to follow a moving target. Unlike classical tracking control, the target is unknown, non-stationary, and its state is revealed sequentially, thus, fitting the framework of online non-stochastic control. We consider the case of quadratic costs and propose a new algorithm, called predictive linear online tracking (PLOT). The algorithm uses recursive least squares with exponential forgetting to learn a time-varying dynamic model of the target. The learned model is used in the optimal policy under the framework of receding horizon control. We show the dynamic regret of PLOT scales with $\mathcal{O}(\sqrt{TV_T})$, where $V_T$ is the total variation of the target dynamics and $T$ is the time horizon. Unlike prior work, our theoretical results hold for non-stationary targets. We implement PLOT on a real quadrotor and provide open-source software, thus, showcasing one of the first successful applications of online control methods on real hardware.

2402.10202

Rylan Schaeffer

Rylan Schaeffer, Nika Zahedi, Mikail Khona, Dhruv Pai, Sang Truong, Yilun Du, Mitchell Ostrow, Sarthak Chandra, Andres Carranza, Ila Rani Fiete, Andrey Gromov, Sanmi Koyejo

Bridging Associative Memory and Probabilistic Modeling

cs.LG

http://creativecommons.org/licenses/by-nc-nd/4.0/

Associative memory and probabilistic modeling are two fundamental topics in artificial intelligence. The first studies recurrent neural networks designed to denoise, complete and retrieve data, whereas the second studies learning and sampling from probability distributions. Based on the observation that associative memory's energy functions can be seen as probabilistic modeling's negative log likelihoods, we build a bridge between the two that enables useful flow of ideas in both directions. We showcase four examples: First, we propose new energy-based models that flexibly adapt their energy functions to new in-context datasets, an approach we term \textit{in-context learning of energy functions}. Second, we propose two new associative memory models: one that dynamically creates new memories as necessitated by the training data using Bayesian nonparametrics, and another that explicitly computes proportional memory assignments using the evidence lower bound. Third, using tools from associative memory, we analytically and numerically characterize the memory capacity of Gaussian kernel density estimators, a widespread tool in probababilistic modeling. Fourth, we study a widespread implementation choice in transformers -- normalization followed by self attention -- to show it performs clustering on the hypersphere. Altogether, this work urges further exchange of useful ideas between these two continents of artificial intelligence.

2402.10986

Gagan Bhatia

Gagan Bhatia, El Moatez Billah Nagoudi, Hasan Cavusoglu, Muhammad Abdul-Mageed

FinTral: A Family of GPT-4 Level Multimodal Financial Large Language Models

cs.CL cs.AI

http://creativecommons.org/licenses/by-sa/4.0/

We introduce FinTral, a suite of state-of-the-art multimodal large language models (LLMs) built upon the Mistral-7b model and tailored for financial analysis. FinTral integrates textual, numerical, tabular, and image data. We enhance FinTral with domain-specific pretraining, instruction fine-tuning, and RLAIF training by exploiting a large collection of textual and visual datasets we curate for this work. We also introduce an extensive benchmark featuring nine tasks and 25 datasets for evaluation, including hallucinations in the financial domain. Our FinTral model trained with direct preference optimization employing advanced Tools and Retrieval methods, dubbed FinTral-DPO-T&R, demonstrates an exceptional zero-shot performance. It outperforms ChatGPT-3.5 in all tasks and surpasses GPT-4 in five out of nine tasks, marking a significant advancement in AI-driven financial technology. We also demonstrate that FinTral has the potential to excel in real-time analysis and decision-making in diverse financial contexts. The GitHub repository for \textit{FinTral} is available at \url{https://github.com/UBC-NLP/fintral}.

2402.11061

Bin Han

Anthony Kiggundu, Bin Han, Dennis Krummacker, and Hans D. Schotten

Chronicles of jockeying in queuing systems

Submitted to ACM Computing Surveys

cs.NI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

The relevance of studies in queuing theory in social systems has inspired its adoption in other mainstream technologies with its application in distributed and communication systems becoming an intense research domain. Considerable work has been done regarding the application of the impatient queuing phenomenon in distributed computing to achieve optimal resource sharing and allocation for performance improvement. Generally, there are two types of common impatient queuing behaviour that have been well studied, namely balking and reneging, respectively. In this survey, we are interested in the third type of impatience: jockeying, a phenomenon that draws origins from impatient customers switching from one queue to another. This survey chronicles classical and latest efforts that labor to model and exploit the jockeying behaviour in queuing systems, with a special focus on those related to information and communication systems, especially in the context of Multi-Access Edge Computing. We comparatively summarize the reviewed literature regarding their methodologies, invoked models, and use cases.

2402.11530

Bo Zhao

Muyang He, Yexin Liu, Boya Wu, Jianhao Yuan, Yueze Wang, Tiejun Huang, Bo Zhao

Efficient Multimodal Learning from Data-centric Perspective

cs.CV

http://creativecommons.org/licenses/by/4.0/

Multimodal Large Language Models (MLLMs) have demonstrated notable capabilities in general visual understanding and reasoning tasks. However, their deployment is hindered by substantial computational costs in both training and inference, limiting accessibility to the broader research and user communities. A straightforward solution is to leverage smaller pre-trained vision and language models, which inevitably cause significant performance drops. In this paper, we demonstrate the possibility of training a smaller but better MLLM with high-quality training data. Specifically, we introduce Bunny, a family of lightweight MLLMs with flexible vision and language backbones for efficient multimodal learning from selected training data. Experiments show that our Bunny-4B/8B outperforms the state-of-the-art large MLLMs on multiple benchmarks. We expect that this work can provide the community with a clean and flexible open-source tool for further research and development. The code, models, and data can be found in https://github.com/BAAI-DCAI/Bunny.