October 2022 arXiv papers — page 113

Andrea Apicella, Francesco Isgrò, Andrea Pollastro, Roberto Prevete

An interesting case of the well-known Dataset Shift Problem is the classification of Electroencephalogram (EEG) signals in the context of Brain-Computer Interface (BCI). The non-stationarity of EEG signals can lead to poor generalisation performance in BCI classification systems used in different sessions, also from the same subject. In this paper, we start

Wentao Zhu, Xiaoxuan Ma, Zhaoyang Liu, Libin Liu

We present a unified perspective on tackling various human-centric video tasks by learning human motion representations from large-scale and heterogeneous data resources. Specifically, we propose a pretraining stage in which a motion encoder is trained to recover the underlying 3D motion from noisy partial 2D observations. The motion representations acquired

Shi-Zeng Lin

Recent experiments [Persky {\it{et al.}}, Nature {\bf{607}}, 692 (2022)] demonstrate a magnetic memory effect in 4Hb-$\mathrm{TaS_2}$ above its superconducting transition temperature, where Abriokosov vortices are generated spontaneously by lowering temperature at zero magnetic field after field training the normal state. Motivated by the experiment, we prop

Paola B. Manasero, Jorge Oviedo

In a many-to-many matching model in which agents' preferences satisfy substitutability and the law of aggregate demand, we proof the General Manipulability Theorem. We result generalizes the presented in Sotomayor (1996 and 2012) for the many-to-one model. In addition, we show General Manipulability Theorem fail when agents' preferences satisfy only substitu

Benjamin Gammage, Justin Hilburn

We propose a conjecture on the categorical trace of the 2-category of perverse schobers (expected to model the Fukaya-Fueter 2-category of a holomorphic symplectic space). By proving a Betti geometric version of Tate's thesis, and combining it with our previous 3d mirror symmetry equivalence and the Ben-Zvi--Nadler--Preygel result on spectral traces, we are

A. J. Cooper, A. Rowlinson, R. A. M. J. Wijers, C. Bassa

We present LOFAR imaging observations from the April/May 2020 active episode of magnetar SGR 1935+2154. We place the earliest radio limits on persistent emission following the low-luminosity fast radio burst FRB 200428 from the magnetar. We also perform an image-plane search for transient emission and find no radio flares during our observations. We examine

Aaron Palmer, Zhiyi Chi, Derek Aguiar, Jinbo Bi

We develop a new type of generative autoencoder called the Goodness-of-Fit Autoencoder (GoFAE), which incorporates GoF tests at two levels. At the minibatch level, it uses GoF test statistics as regularization objectives. At a more global level, it selects a regularization coefficient based on higher criticism, i.e., a test on the uniformity of the local GoF

Enric Boix-Adsera, Hannah Lawrence, George Stepaniants, Philippe Rigollet

Comparing the representations learned by different neural networks has recently emerged as a key tool to understand various architectures and ultimately optimize them. In this work, we introduce GULP, a family of distance measures between representations that is explicitly motivated by downstream predictive tasks. By construction, GULP provides uniform contr

Stanislav K. Ignatov, Artëm E. Masunov

The electronic parameters and, in particular, the isotropic electrostatic polarizability (IEP) of sub-nanoscale magnesium clusters were studied in an expanded set of 1237 structurally unique isomers found in the course of direct global DFT optimization of the structure of Mg2-Mg32 clusters at the BP86 level, as well as using global optimization based on DFT-

Letif Mones, Sean Lovett

Convergence (virtual) bidding is an important part of two-settlement electric power markets as it can effectively reduce discrepancies between the day-ahead and real-time markets. Consequently, there is extensive research into the bidding strategies of virtual participants aiming to obtain optimal bids to submit to the day-ahead market. In this paper, we int

Anna Harvey, Orion Taylor

We leverage new data on daily individual-level jail records and exploit the timing of incarceration to estimate the causal effects of jail incarceration on voting from jail in 2020. We find that registered voters booked into county jails for the full duration of 2020 voting days were on average 46% less likely to vote in 2020, relative to registered voters b

Steve Kench, Isaac Squires, Amir Dahari, Samuel J Cooper

3D microstructural datasets are commonly used to define the geometrical domains used in finite element modelling. This has proven a useful tool for understanding how complex material systems behave under applied stresses, temperatures and chemical conditions. However, 3D imaging of materials is challenging for a number of reasons, including limited field of

Neelu Jyoti Ahuja, Adarsh Kumar, Monika Thapliyal, Sarthika Dutt

Underwater drones have found a place in oceanography, oceanic research, bathymetric surveys, military, surveillance, monitoring, undersea exploration, mining, commercial diving, photography and several other activities. Drones housed with several sensors and complex propulsion systems help oceanographic scientists and undersea explorers to map the seabed, st

Andrew M. Childs, Tongyang Li, Jin-Peng Liu, Chunhao Wang

Given a convex function $f\colon\mathbb{R}^{d}\to\mathbb{R}$, the problem of sampling from a distribution $\propto e^{-f(x)}$ is called log-concave sampling. This task has wide applications in machine learning, physics, statistics, etc. In this work, we develop quantum algorithms for sampling log-concave distributions and for estimating their normalizing con

Lyan Abdul, Jocelyn Xu, Alexander Sotra, Abbas Chaudary

Stem cell-derived organoids are a promising tool to model native human tissues as they resemble human organs functionally and structurally compared to traditional monolayer cell-based assays. For instance, colon organoids can spontaneously develop crypt-like structures similar to those found in the native colon. While analyzing the structural development of

Christopher Morency, Daniel J. Stilwell

We seek to rigorously evaluate the benefit of using a few beams rather than a single beam for a low-cost obstacle avoidance sonar for small AUVs. For a small low-cost AUV, the complexity, cost, and volume required for a multi-beam forward looking sonar are prohibitive. In contrast, a single-beam system is relatively easy to integrate into a small AUV, but do

Yiming Huo, Xiaodai Dong, Nuwan Ferdinand

With the fifth-generation (5G) networks widely commercialized and fast deployed, the sixth-generation (6G) wireless communication is envisioned to provide competitive quality of service (QoS) in multiple aspects to global users. The critical and underlying research of the 6G is, firstly, highly dependent on the precise modeling and characterization of the wi

Christopher Morency, Daniel J. Stilwell, Sebastian Hess

This paper describes a high-fidelity sonar model and a simulation environment that implements the model. The model and simulation environment have been developed to aid in the design of a forward looking sonar for autonomous underwater vehicles (AUVs). The simulator achieves real-time visualization through ray tracing and approximation. The simulator facilit

Rabindranath Bag, Sachith E. Dissanayake, Han Yan, Zhenzhong Shi

Recently a new class of quantum magnets, the so-called breathing pyrochlore spin systems, have attracted much attention due to their potential to host exotic emergent phenomena. Here, we present magnetometry, heat capacity, thermal conductivity, Muon-spin relaxation, and polarized inelastic neutron scattering measurements performed on high-quality single-cry

Stephen Majeski, Hantao Ji

We investigate the Fermi acceleration of charged particles in 2D MHD anti-parallel plasmoid reconnection, finding a drastic enhancement in energization rate $\dot{\varepsilon}$ over a standard Fermi model of $\dot{\varepsilon} \sim \varepsilon$. The shrinking particle orbit width around a magnetic island due to $\vec{E}\times\vec{B}$ drift produces a $\dot{\

Ugo Bindini, Guy Bouchitté

Optimization problems on probability measures in $\mathbb{R}^d$ are considered where the cost functional involves multi-marginal optimal transport. In a model of $N$ interacting particles, like in Density Functional Theory, the interaction cost is repulsive and described by a two-point function $c(x,y) =\ell(|x-y|)$ where $\ell: \mathbb{R}_+ \to [0,\infty]$

Manuel E. Rodrigues, Henrique A. Vieira

In this paper we present a Bardeen solution surrounded by a cloud of strings fluid. We show how this model has the same event horizon characteristic as the Bardeen solution, however, the parameter of the strings make the solution singular at the origin. We also analyze the solution from a thermodynamic point of view. We calculate the system's state function,

Hamid Abchir, Soukaina Lamsifer

We improve the lower bound for the minimum number of colors for linear Alexander quandle colorings of a knot given in Theorem 1.2 of Colorings beyond Fox: The other linear Alexander quandles (Linear Algebra and its Applications, Vol. 548, 2018). We express this lower bound in terms of the degree k of the reduced Alexander polynomial of the considered knot. W

Anjith George, Amir Mohammadi, Sebastien Marcel

Heterogeneous Face Recognition (HFR) refers to matching face images captured in different domains, such as thermal to visible images (VIS), sketches to visible images, near-infrared to visible, and so on. This is particularly useful in matching visible spectrum images to images captured from other modalities. Though highly useful, HFR is challenging because

Vijay S. Mahadevan, David Lenz, Iulian Grindeanu, Thomas Peterka

Compactly expressing large-scale datasets through Multivariate Functional Approximations (MFA) can be critically important for analysis and visualization to drive scientific discovery. Tackling such problems requires scalable data partitioning approaches to compute MFA representations in amenable wall clock times. We introduce a fully parallel scheme to redu

Kostov Blechin, Alvarez-Esteban Ramón, Bécue-Bertaut Mónica, Husson François

This paper focuses on the analysis of open-ended questions answered in different languages. Closed-ended questions, called contextual variables, are asked to all respondents in order to understand the relationships between the free and the closed responses among the different samples since the latter assumably affect the word choices. We have developed "Mult

Sergei V. Kalinin, Rama Vasudevan, Yongtao Liu, Ayana Ghosh

We pose that microscopy offers an ideal real-world experimental environment for the development and deployment of active Bayesian and reinforcement learning methods. Indeed, the tremendous progress achieved by machine learning (ML) and artificial intelligence over the last decade has been largely achieved via the utilization of static data sets, from the par

Francois Meyer, Jan Buys

Subwords have become the standard units of text in NLP, enabling efficient open-vocabulary models. With algorithms like byte-pair encoding (BPE), subword segmentation is viewed as a preprocessing step applied to the corpus before training. This can lead to sub-optimal segmentations for low-resource languages with complex morphologies. We propose a subword se

Maciej Hendzel, Józef Spałek

We analyze two-particle binding factors for the case of \ch{H2} molecule with the help of our original Exact Diagonalization \textit{Ab Intio} (EDABI) approach. Explicitly, we redefine the many-particle covalency and ionicity factors as a function of interatomic distance. Insufficiency of those basic characteristics is stressed and the concept of \textit{ato

Jonathan Frassineti, Pietro Bonfà, Giuseppe Allodi, Erick Garcia

The recently discovered vanadium-based kagome metals AV$_3$Sb$_5$ (A = K, Rb, Cs) offer the possibility to study the interplay between competing electronic orderings, such as charge density order and superconductivity. We focus on the former and provide a comprehensive set of $^{51}$V, $^{87}$Rb, and $^{121}$Sb magnetic resonance measurements on an RbV$_3$Sb

Fengwu Sun, Eiichi Egami

We analysed the high-resolution (up to $\sim$0.2") ALMA CO (2-1) and 1.3 mm dust continuum data of eight gas-rich post-starburst galaxies (PSBs) in the local Universe, six of which had been studied by Smercina et al. (2022). In contrast to this study reporting the detections of extraordinarily compact (i.e., unresolved) reservoirs of molecular gas in the six

Colin Vendromin, Marc M. Dignam

We prove that the density operator for the nonlinearly-generated quantum state of light in the $M$ lossy nonorthogonal quasimodes of a nanocavity system has the analytic form of a multimode squeezed thermal state, where the time-dependence of the squeezing and thermal photon parameters are given by a set of $3M$ coupled differential equations. We apply our a

Sylvain Moinard, Edouard Oudet, Didier Piau, Eric Coissac

Modelling in biology must adapt to increasingly complex and massive data. The efficiency of the inference algorithms used to estimate model parameters is therefore questioned. Many of these are based on stochastic optimization processes that require significant computing time. We introduce the Fixed Landscape Inference MethOd (flimo), a new likelihood-free i

João Barata, Andrey V. Sadofyev, Xin-Nian Wang

We study gradient corrections to the transport equation for energetic light partons in dense QCD environments. In the diffusion limit, the transport dynamics is solely controlled by small-angle elastic scatterings, leading to transverse momentum broadening with respect to the parton's initial direction. Such a parton propagation is usually considered in the

Qinqing Zheng, Mikael Henaff, Brandon Amos, Aditya Grover

Natural agents can effectively learn from multiple data sources that differ in size, quality, and types of measurements. We study this heterogeneity in the context of offline reinforcement learning (RL) by introducing a new, practically motivated semi-supervised setting. Here, an agent has access to two sets of trajectories: labelled trajectories containing

Martin Doubek, Branislav Jurčo, Lada Peksová, Ján Pulmann

Modular operads relevant to string theory can be equipped with an additional structure, coming from the connected sum of surfaces. Motivated by this example, we introduce a notion of connected sum for general modular operads. We show that a connected sum induces a commutative product on the space of functions associated to the modular operad. Moreover, we co

Yi Zeng, Minzhou Pan, Himanshu Jahagirdar, Ming Jin

Given the volume of data needed to train modern machine learning models, external suppliers are increasingly used. However, incorporating external data poses data poisoning risks, wherein attackers manipulate their data to degrade model utility or integrity. Most poisoning defenses presume access to a set of clean data (or base set). While this assumption ha

Ruifeng Xie, Max E. Austin, Kenneth W. Gentle, Craig C. Petty

DIII-D discharges that transition to H-mode solely with off-axis electron cyclotron heating (ECH) often exhibit strong off-axis peaking of electron temperature profiles at the heating location. Electron heat transport properties near these off-axis temperature peaks have been studied using modulated ECH. The Fourier analyzed electron temperature data have be

Jose M. Diego, Ashish K. Meena, Nathan J. Adams, Tom Broadhurst

The first JWST data on the massive colliding cluster El Gordo confirm 23 known families of multiply lensed images and identify 8 new members of these families. Based on these families, which have been confirmed spectroscopically by MUSE, we derived an initial lens model. This model guided the identification of 37 additional families of multiply lensed galaxi

Aristomenis Donos, Polydoros Kailidis

We study the nearly critical behaviour of holographic superfluids at finite temperature and chemical potential in their probe limit. This allows us to examine the coupled dynamics of the full complex order parameter with the charge density of the system. We derive an effective theory for the long wavelength limit of the gapless and pseudo-gapped modes by usi

Zhenyu Yang, Kyle Lafata, Eugene Vaios, Zongsheng Hu

The projection of planar MRI data onto a spherical surface is equivalent to a nonlinear image transformation that retains global anatomical information. By incorporating this image transformation process in our proposed spherical projection-based U-Net (SPU-Net) segmentation model design, multiple independent segmentation predictions can be obtained from a s

Fredrik Hellström, Giuseppe Durisi

Recent work has established that the conditional mutual information (CMI) framework of Steinke and Zakynthinou (2020) is expressive enough to capture generalization guarantees in terms of algorithmic stability, VC dimension, and related complexity measures for conventional learning (Harutyunyan et al., 2021, Haghifam et al., 2021). Hence, it provides a unifi

Stephen T. Krauss, Daniel J. Stilwell

In this work, we present an aided inertial navigation system for an autonomous underwater vehicle (AUV) using an unscented Kalman filter on manifolds (UKF-M). The inertial navigation estimate is aided by a Doppler velocity log (DVL), depth sensor, acoustic range and, while on the surface, GPS. The sensor model for each navigation sensor on the AUV is explici

Di Tang, Rui Zhu, XiaoFeng Wang, Haixu Tang

With extensive studies on backdoor attack and detection, still fundamental questions are left unanswered regarding the limits in the adversary's capability to attack and the defender's capability to detect. We believe that answers to these questions can be found through an in-depth understanding of the relations between the primary task that a benign model i

S. Mukhopadhyay, J. Senior, J. Saez-Mollejo, D. Puglia

Quantum phase transitions typically result in a broadened critical or crossover region at nonzero temperature. Josephson arrays are a model of this phenomenon, exhibiting a superconductor-insulator transition at a critical wave impedance, and a well-understood insulating phase. Yet high-impedance arrays used in quantum computing and metrology apparently evad

Pinyan Lu, Enze Sun, Chenghan Zhou

Submodular over signal (SOS) defines a family of interesting functions for which there exist truthful mechanisms with constant approximation to the social welfare for agents with interdependent valuations. The best-known truthful auction is of $4$-approximation and a lower bound of 2 was proved. We propose a new and simple truthful mechanism to achieve an ap

Canxun Zhang, Tiancong Zhu, Tomohiro Soejima, Salman Kahn

In recent years, correlated insulating states, unconventional superconductivity, and topologically non-trivial phases have all been observed in several moir\'e heterostructures. However, understanding of the physical mechanisms behind these phenomena is hampered by the lack of local electronic structure data. Here, we use scanning tunnelling microscopy and s

V. Denysiuk

Methods of constructing trigonometric fundamental splines with constant sign and sign-changing convergence factors are given. An example and graphics illustrating the concepts of convergence and interpolation grids are given. Some methods of constructing constant-sign and sign-changing coefficients of convergence of trigonometric splines are considered.

A. C. Trapp, Steven R. Furlanetto, Frederick B. Davies

Lyman-alpha emitters (LAEs) are excellent probes of the reionization process, as they must be surrounded by large ionized bubbles in order to be visible during the reionization era. Large ionized regions are thought to correspond to over-dense regions and may be protoclusters, making them interesting test-beds for early massive structures. Close associations

Shifra Mandel, Michael M. Shara, David Zurek, Charlie Conroy

Accurate determination of the rates of nova eruptions in different kinds of galaxies give us strong constraints on those galaxies' underlying white dwarf and binary populations, and those stars' spatial distributions. Until 2016, limitations inherent in ground-based surveys of external galaxies - and dust extinction in the Milky Way - significantly hampered

Bassel Heiba Elfeky, Joseph J. Cuozzo, Neda Lotfizadeh, William F. Schiela

In the presence of a 4$\pi$-periodic contribution to the current phase relation, for example in topological Josephson junctions, odd Shapiro steps are expected to be missing. While missing odd Shapiro steps have been observed in several material systems and interpreted in the context of topological superconductivity, they have also been observed in topologic

Yaser Souri, Yazan Abu Farha, Emad Bahrami, Gianpiero Francesca

Action segmentation is the task of predicting an action label for each frame of an untrimmed video. As obtaining annotations to train an approach for action segmentation in a fully supervised way is expensive, various approaches have been proposed to train action segmentation models using different forms of weak supervision, e.g., action transcripts, action

Marco Calzà, João G. Rosa

We propose a method to determine the mass and spin of primordial black holes (PBHs) in the mass range $5\times 10^7-10^{12}$ kg (Hawking temperatures $\sim10$ MeV $-200$ GeV), based on measuring the energy of specific features in the photon Hawking emission spectrum, including both primary and secondary components. This is motivated by scenarios where PBHs i

Griffin Hosseinzadeh, David J. Sand, Jacob E. Jencson, Jennifer E. Andrews

We present near- and mid-infrared (0.9-18 $\mu$m) photometry of supernova (SN) 2021afdx, which was imaged serendipitously with the James Webb Space Telescope (JWST) as part of its Early Release Observations of the Cartwheel Galaxy. Our ground-based optical observations show it is likely to be a Type IIb SN, the explosion of a yellow supergiant, and its infra

Elham Rahimi, Evan Vienneau, Nassim Bozorgnia, Andrew Robertson

We study the effects of dark matter self-interactions on the local dark matter distribution in selected Milky Way-like galaxies in the EAGLE hydrodynamical simulations. The simulations were run with two different self-interacting dark matter models, a constant and velocity-dependent self-interaction cross-section. We find that the local dark matter velocity

Monica Jinwoo Kang, Craig Lawrie, Ki-Hong Lee, Jaewon Song

We study a rich set of four-dimensional superconformal field theories (SCFTs) with both central charges identical: $a = c$. These are constructed via the diagonal $\mathcal{N}=2$ or $\mathcal{N}=1$ gauging of the flavor symmetry $G$ of a collection of $\mathcal{N}=2$ Argyres-Douglas theories of type $\mathcal{D}_p(G)$, with or without adjoint chiral multiple

Rui Ribeiro, Luísa Coheur

In this paper, we investigate the problem of including relevant information as context in open-domain dialogue systems. Most models struggle to identify and incorporate important knowledge from dialogues and simply use the entire turns as context, which increases the size of the input fed to the model with unnecessary information. Additionally, due to the in

Marius de Leeuw, Rafael I. Nepomechie, Ana L. Retore

We introduce new classes of integrable models that exhibit a structure similar to that of flag vector spaces. We present their Hamiltonians, R-matrices and Bethe-ansatz solutions. These models have a new type of generalized graded algebra symmetry.

Yongchan Yoo, Christopher David White, Brian Swingle

We use non-equilibrium steady states to study the effect of dissipation-assisted operator evolution (DAOE) on the scaling behavior of transport in one-dimensional spin chains. We consider three models in the XXZ family: the XXZ model with staggered anisotropy, which is chaotic; XXZ model with no external field and tunable interaction, which is Bethe ansatz i

Brent Tan, S. Peng Oh, Max Gronke

Understanding the survival, growth and dynamics of cold gas is fundamental to galaxy formation. While there has been a plethora of work on `wind tunnel' simulations that study such cold gas in winds, the infall of this gas under gravity is at least equally important, and fundamentally different since cold gas can never entrain. Instead, velocity shear increa

K. D. Neumann, T. W. -S. Holoien, C. S. Kochanek, K. Z. Stanek

We catalog the 443 bright supernovae discovered by the All-Sky Automated Survey for Supernovae (ASAS-SN) in $2018-2020$ along with the 519 supernovae recovered by ASAS-SN and 516 additional $m_{peak}\leq18$ mag supernovae missed by ASAS-SN. Our statistical analysis focuses primarily on the 984 supernovae discovered or recovered in ASAS-SN $g$-band observatio

Cole Johnston, Andrew Tkachenko, Timothy Van Reeth, Dominic M. Bowman

Context. Unambiguous examples of the influence of tides on self-excited, free stellar pulsations have recently been observationally detected in space-based photometric data. Aims. We aim to investigate U Gru and contextualise it within the growing class of tidally influenced pulsators. Initial analysis of U Gru revealed frequencies spaced by the orbital freq

Kim V. Berghaus, Angelo Esposito, Rouven Essig, Mukul Sholapurkar

Dark matter scattering off a nucleus has a small probability of inducing an observable ionization through the inelastic excitation of an electron, called the Migdal effect. We use an effective field theory to extend the computation of the Migdal effect in semiconductors to regions of small momentum transfer to the nucleus, where the final state of the nucleu

Bhavik Kumar, Philipp Hauke, Jad C. Halimeh

In the current drive to quantum-simulate evermore complex gauge-theory phenomena, it is necessary to devise schemes allowing for the control and suppression of unavoidable gauge-breaking errors on different experimental platforms. Although there have been several successful approaches to tackle coherent errors, comparatively little has been done in the way o

Scott Armstrong, Tuomo Kuusi

We give a self-contained introduction to the theory of elliptic homogenization for random coefficient fields, starting from classical qualitative homogenization. The presentation also contains new results, such as optimal estimates (both in terms of stochastic moments and scaling of the error) for coefficient fields which are local functions of Gaussian rand

Gordan Krnjaic, Duncan Rocha, Anastasia Sokolenko

We study a simple model of vector dark matter that couples to Standard Model particles via magnetic dipole interactions. In this scenario, the cosmological abundance arises through the freeze-in mechanism and depends on the dipole coupling, the vector mass, and the reheat temperature. To ensure cosmological metastability, the vector must be lighter than the

D. M. Rowan, T. Jayasinghe, K. Z. Stanek, C. S. Kochanek

Detached eclipsing binaries are the primary tool used to measure precise masses and radii of stars. In our previous paper estimating the parameters of more than 30,000 detached eclipsing binaries, we identified 766 eclipsing binaries with additional features in their All-Sky Automated Survey for Supernovae (ASAS-SN) and Transiting Exoplanet Survey Satellite

Yimeng Tang, Bojun Tao, Hong-Xin Zhang, Guangwen Chen

NGC 2915 is a unique nearby galaxy that is classified as an isolated blue compact dwarf based on its optical appearance but has an extremely extended H i gas disk with prominent Sd-type spiral arms. To unveil the starburst-triggering mystery of NGC 2915, we performed a comprehensive analysis of deep VLT/MUSE integral field spectroscopic observations that cov

Lennart Bittel, Jens Watty, Martin Kliesch

Many optimization methods for training variational quantum algorithms are based on estimating gradients of the cost function. Due to the statistical nature of quantum measurements, this estimation requires many circuit evaluations, which is a crucial bottleneck of the whole approach. We propose a new gradient estimation method to mitigate this measurement ch

Ibrahima Bah, Pierre Heidmann

We construct large classes of non-BPS smooth horizonless geometries that are asymptotic to AdS$_3\times$S$^3\times$T$^4$ in type IIB supergravity. These geometries are supported by electromagnetic flux corresponding to D1-D5 charges. We show that Einstein equations for systems with eight commuting Killing vectors decompose into a set of Ernst equations, ther

Danial Langeroodi, Jens Hjorth

The mid-infrared spectra of star-forming galaxies (SFGs) are characterized by characteristic broad PAH emission features at 3-20 $\mu$m. As these features are redshifted, they are predicted to dominate the flux at specific mid-infrared wavelengths, leading to substantial redshift-dependent color variations in broad-band photometry. The advent of JWST for the

Jeff A. Dror, Stefania Gori, Jacob M. Leedom, Nicholas L. Rodd

A leading direction in the hunt for axion dark matter is to search for its influence on nuclear spins. The detection scheme involves polarizing a sample of nuclei within a strong static magnetic field and then looking for a spin precession induced by the oscillating axion field. We study the axion signal and background contributions that arise in such experi

Yunxiang Liao, Victor Galitski

We use field-theoretic methods to explore the statistics of eigenfunctions of the Floquet operator for a large family of Floquet random quantum circuits. The correlation function of the quasienergy eigenstates is calculated and shown to exhibit random matrix circular unitary ensemble statistics, which is consistent with the analogue of Berry's conjecture for

Jochem Loedeman, Maarten C. Stol, Tengda Han, Yuki M. Asano

With the introduction of the transformer architecture in computer vision, increasing model scale has been demonstrated as a clear path to achieving performance and robustness gains. However, with model parameter counts reaching the billions, classical finetuning approaches are becoming increasingly limiting and even unfeasible when models become hosted as in

Yue Wu, Yu Deng, Jiaolong Yang, Fangyun Wei

Although 2D generative models have made great progress in face image generation and animation, they often suffer from undesirable artifacts such as 3D inconsistency when rendering images from different camera viewpoints. This prevents them from synthesizing video animations indistinguishable from real ones. Recently, 3D-aware GANs extend 2D GANs for explicit

Alex Boyd, Sam Showalter, Stephan Mandt, Padhraic Smyth

In reasoning about sequential events it is natural to pose probabilistic queries such as "when will event A occur next" or "what is the probability of A occurring before B", with applications in areas such as user modeling, medicine, and finance. However, with machine learning shifting towards neural autoregressive models such as RNNs and transformers, proba

Sridhar Pandian Arunachalam, Irmak Güzey, Soumith Chintala, Lerrel Pinto

A fundamental challenge in teaching robots is to provide an effective interface for human teachers to demonstrate useful skills to a robot. This challenge is exacerbated in dexterous manipulation, where teaching high-dimensional, contact-rich behaviors often require esoteric teleoperation tools. In this work, we present Holo-Dex, a framework for dexterous ma

Vincent Tao Hu, David W Zhang, Yuki M. Asano, Gertjan J. Burghouts

Diffusion models have demonstrated remarkable progress in image generation quality, especially when guidance is used to control the generative process. However, guidance requires a large amount of image-annotation pairs for training and is thus dependent on their availability, correctness and unbiasedness. In this paper, we eliminate the need for such annota

Juri Opitz, Anette Frank

Recently, astonishing advances have been observed in AMR parsing, as measured by the structural Smatch metric. In fact, today's systems achieve performance levels that seem to surpass estimates of human inter annotator agreement (IAA). Therefore, it is unclear how well Smatch (still) relates to human estimates of parse quality, as in this situation potential

Gaoyue Zhou, Liyiming Ke, Siddhartha Srinivasa, Abhinav Gupta

Offline reinforcement learning (ORL) holds great promise for robot learning due to its ability to learn from arbitrary pre-generated experience. However, current ORL benchmarks are almost entirely in simulation and utilize contrived datasets like replay buffers of online RL agents or sub-optimal trajectories, and thus hold limited relevance for real-world ro

Yijun Zuo

The least trimmed squares (LTS) estimator is popular in location, regression, machine learning, and AI literature. Despite the empirical version of least trimmed squares (LTS) being repeatedly studied in the literature, the population version of the LTS has never been introduced and studied. The lack of the population version hinders the study of the large s

Kelly Ramsay, Aukosh Jagannath, Shoja'eddin Chenouri

Statistical tools which satisfy rigorous privacy guarantees are necessary for modern data analysis. It is well-known that robustness against contamination is linked to differential privacy. Despite this fact, using multivariate medians for differentially private and robust multivariate location estimation has not been systematically studied. We develop novel

Chaofei Wang, Qisen Yang, Rui Huang, Shiji Song

Knowledge distillation is an effective approach to learn compact models (students) with the supervision of large and strong models (teachers). As empirically there exists a strong correlation between the performance of teacher and student models, it is commonly believed that a high performing teacher is preferred. Consequently, practitioners tend to use a we

Tanmay Khale

We establish the first explicit form of the Vinogradov--Korobov zero-free region for Dirichlet $L$-functions.

Nelson F. Liu, Ananya Kumar, Percy Liang, Robin Jia

Recent results in image classification and extractive question answering have observed that pre-trained models trained on less in-distribution data have better out-of-distribution performance. However, it is unclear how broadly these trends hold. We conduct a large empirical study across three tasks, three broadly-applicable modeling interventions (increasin

Han Xiao, Wenzhao Zheng, Zheng Zhu, Jie Zhou

Data mixing strategies (e.g., CutMix) have shown the ability to greatly improve the performance of convolutional neural networks (CNNs). They mix two images as inputs for training and assign them with a mixed label with the same ratio. While they are shown effective for vision transformers (ViTs), we identify a token fluctuation phenomenon that has suppresse

Atul Singh Arora, Andrea Coladangelo, Matthew Coudron, Alexandru Gheorghiu

We give a comprehensive characterization of the computational power of shallow quantum circuits combined with classical computation. Specifically, for classes of search problems, we show that the following statements hold, relative to a random oracle: (a) $\mathsf{BPP}^{\mathsf{QNC}^{\mathsf{BPP}}} \neq \mathsf{BQP}$. This refutes Jozsa's conjecture [QIP 05]

Christopher W. Lynn, Caroline M. Holmes, Stephanie E. Palmer

Many complex systems--from social and communication networks to biological networks and the Internet--are thought to exhibit scale-free structure. However, prevailing explanations rely on the constant addition of new nodes, an assumption that fails dramatically in some real-world settings. Here, we propose a model in which nodes are allowed to die, and their

Ryan M. Lau, Matthew J. Hankins, Yinuo Han, Ioannis Argyriou

Massive colliding-wind binaries that host a Wolf-Rayet (WR) star present a potentially important source of dust and chemical enrichment in the interstellar medium (ISM). However, the chemical composition and survival of dust formed from such systems is not well understood. The carbon-rich WR (WC) binary WR~140 presents an ideal astrophysical laboratory for i

Gustav Uhre Jakobsen, Gustav Mogull

Using the spinning, supersymmetric Worldline Quantum Field Theory formalism we compute the momentum impulse and spin kick from a scattering of two spinning black holes or neutron stars up to quadratic order in spin at third post-Minkowskian (PM) order, including radiation-reaction effects and with arbitrarily mis-aligned spin directions. Parts of these obser

Kyle Lee, James Mulligan, Mateusz Płoskoń, Felix Ringer

We explore machine learning-based jet and event identification at the future Electron-Ion Collider (EIC). We study the effectiveness of machine learning-based classifiers at relatively low EIC energies, focusing on (i) identifying the flavor of the jet and (ii) identifying the underlying hard process of the event. We propose applications of our machine learn

M. Tokarchuk

Based on a chain of BBGKI equations with a modified boundary condition that takes into account multiparticle correlations, kinetic equations in the approximate "pairs" collisions and in the polarization approximation, taking into account the interaction through the third particle, obtained. The specifics of the model representation of the pair potential of p

Kenta Takatsu, Ted Westling

In this article, we study nonparametric inference for a covariate-adjusted regression function. This parameter captures the average association between a continuous exposure and an outcome after adjusting for other covariates. In particular, under certain causal conditions, this parameter corresponds to the average outcome had all units been assigned to a sp

Ruqi Zhang, Qiang Liu, Xin T. Tong

Sampling methods, as important inference and learning techniques, are typically designed for unconstrained domains. However, constraints are ubiquitous in machine learning problems, such as those on safety, fairness, robustness, and many other properties that must be satisfied to apply sampling results in real-life applications. Enforcing these constraints o

Haina Wang, Salvatore Torquato

The Zhang-Torquato conjecture [Phys. Rev. E 101, 032124 (2020)] states that any realizable pair correlation function $g_2({\bf r})$ or structure factor $S({\bf k})$ of a translationally invariant nonequilibrium system can be attained by an equilibrium ensemble involving only (up to) effective two-body interactions. To test this conjecture, we consider two si

Mahendra K. Verma, Soumyadeep Chatterjee

Using numerical simulations, we show that the asymptotic states of two-dimensional (2D) Euler turbulence exhibit large-scale flow structures due to nonzero energy transfers among small wavenumber modes. These asymptotic states, which depend on the initial conditions, are out of equilibrium, and they are different from the predictions of Onsager and Kraichnan

Janvijay Singh, Fan Bai, Zhen Wang

Cross-task knowledge transfer via multi-task learning has recently made remarkable progress in general NLP tasks. However, entity tracking on the procedural text has not benefited from such knowledge transfer because of its distinct formulation, i.e., tracking the event flow while following structural constraints. State-of-the-art entity tracking approaches

Lorenzo Bonicelli, Matteo Boschini, Angelo Porrello, Concetto Spampinato

Rehearsal approaches enjoy immense popularity with Continual Learning (CL) practitioners. These methods collect samples from previously encountered data distributions in a small memory buffer; subsequently, they repeatedly optimize on the latter to prevent catastrophic forgetting. This work draws attention to a hidden pitfall of this widespread practice: rep

Zhangdie Yuan, Songbo Hu, Ivan Vulić, Anna Korhonen

Acquiring factual knowledge with Pretrained Language Models (PLMs) has attracted increasing attention, showing promising performance in many knowledge-intensive tasks. Their good performance has led the community to believe that the models do possess a modicum of reasoning competence rather than merely memorising the knowledge. In this paper, we conduct a co

Jonas Geiping, Micah Goldblum, Gowthami Somepalli, Ravid Shwartz-Ziv

Despite the clear performance benefits of data augmentations, little is known about why they are so effective. In this paper, we disentangle several key mechanisms through which data augmentations operate. Establishing an exchange rate between augmented and additional real data, we find that in out-of-distribution testing scenarios, augmentations which yield