May 2024 arXiv papers

Austin Hoover, Jonathan C. Wong

Particle accelerators generate charged-particle beams with tailored distributions in six-dimensional position-momentum space (phase space). Knowledge of the phase space distribution enables model-based beam optimization and control. In the absence of direct measurements, the distribution must be tomographically reconstructed from its projections. In this pap

LHCb collaboration, R. Aaij, A. S. W. Abdelmotteleb, C. Abellan Beteta

A search for radiative decay of $B^0_s$ mesons to orbitally excited $K^+K^-$ states is performed using proton proton collisions recorded by the \mbox{LHCb}\xspace experiment, corresponding to an integrated luminosity of 9~fb$^{-1}$. The dikaon spectrum in the mass range $m_{KK}<2400$~{\ensuremath{\,\text{Me\kern -0.1em V\!/}c^2}\xspace} is dominated by the $

De-Zhang Li, Xin Wang, Xiao-Bao Yang

Ising model is famous in condensed matter and statistical physics. In this work we present a free-fermion formulation of the two-dimensional classical Ising models on the honeycomb, triangular and Kagomé lattices. Each Ising model is studied in the cases of a zero field and of an imaginary field $i(π/2){k_B}T$. We employ the decorated lattice technique, star

Yifan Zeng, Ojas Tendolkar, Raymond Baartmans, Qingyun Wu

Ranking passages by prompting a large language model (LLM) can achieve promising performance in modern information retrieval (IR) systems. A common approach to sort the ranking list is by prompting LLMs for a pairwise or setwise comparison which often relies on sorting algorithms. However, sorting-based methods require consistent comparisons to correctly sor

Caucher Birkar, Jia Jia, Artan Sheshmani

In this paper we build bridges between moduli theory of sheaf stable pairs on one hand and birational geometry on the other hand. We will in particular treat moduli of sheaf stable pairs on smooth projective curves in detail and present some calculations in low degrees. We will also outline problems in various directions.

William Hogan, Jingbo Shang

Recent research efforts have explored the potential of leveraging natural language inference (NLI) techniques to enhance relation extraction (RE). In this vein, we introduce MetaEntailRE, a novel adaptation method that harnesses NLI principles to enhance RE performance. Our approach follows past works by verbalizing relation classes into class-indicative hyp

Maximillian Chen, Ruoxi Sun, Tomas Pfister, Sercan Ö. Arık

Large language models (LLMs), optimized through human feedback, have rapidly emerged as a leading paradigm for developing intelligent conversational assistants. However, despite their strong performance across many benchmarks, LLM-based agents might still lack conversational skills such as disambiguation -- when they are faced with ambiguity, they often over

Gonzalo Navarro, Alejandro Pacheco

We introduce a data structure for counting pattern occurrences in texts compressed with any run-length context-free grammar. Our structure uses space proportional to the grammar size and counts the occurrences of a pattern of length $m$ in a text of length $n$ in time \(O(m\log^{2+ε} n)\), for any constant \(ε> 0\) chosen at indexing time. This is the first

Neelam Shukla, Artem G. Volosniev, Jeremy R. Armstrong

We study a modified three-dimensional Gross-Pitaevski equation that describes a static impurity in a dipolar Bose-Einstein condensate (BEC). Our focus is on the interplay between the shape of the impurity and the anisotropy of the medium manifested in the energy and the density of the system. Without external confinement, properties of the system are derived

Michail Mamalakis, Héloïse de Vareilles, Graham Murray, Pietro Lio

Explainability is a critical factor in enhancing the trustworthiness and acceptance of artificial intelligence (AI) in healthcare, where decisions directly impact patient outcomes. Despite advancements in AI interpretability, clear guidelines on when and to what extent explanations are required in medical applications remain lacking. We propose a novel categ

Jie JW Wu, Fatemeh H Fard

Large language models (LLMs) have significantly improved their ability to perform tasks in the field of code generation. However, there is still a gap between LLMs being capable coders and being top-tier software engineers. Based on the observation that top-level software engineers often ask clarifying questions to reduce ambiguity in both requirements and c

Peter Dotti, Yifei Bai, Toshihiko Shimasaki, Anna R. Dardia

The interplay of various localizing mechanisms is a central topic of modern condensed matter physics. In this work we experimentally explore the interplay between quasiperiodic disorder and periodic driving, each of which in isolation is capable of driving a metal-insulator phase transition. Using a 1D quasiperiodic cold-atom chain we measure transport acros

Chen Feng, Duolikun Danier, Fan Zhang, Alex Mackin

Visual artifacts are often introduced into streamed video content, due to prevailing conditions during content production and delivery. Since these can degrade the quality of the user&#39;s experience, it is important to automatically and accurately detect them in order to enable effective quality measurement and enhancement. Existing detection methods often

Jinchao Zhu, Yuxuan Wang, Siyuan Pan, Pengfei Wan

The Stable Diffusion Model (SDM) is a prevalent and effective model for text-to-image (T2I) and image-to-image (I2I) generation. Despite various attempts at sampler optimization, model distillation, and network quantification, these approaches typically maintain the original network architecture. The extensive parameter scale and substantial computational de

John T. Halloran, Manbir Gulati, Paul F. Roysdon

Mamba state-space models (SSMs) have recently outperformed state-of-the-art (SOTA) Transformer large language models (LLMs) in various tasks and been widely adapted. However, a major concern for stable learning in recurrent-based deep models (such as SSMs) is the sensitivity of their recurrent dynamics. Despite widespread adaptation, the sensitivity of Mamba

S. Y. Lou

The quest to reveal the physical essence of the infinitely many symmetries and conservation laws that are intrinsic to integrable systems has historically posed a significant challenge at the confluence of physics and mathematics. This scholarly investigation delves into five open problems related to these boundless symmetries within integrable systems by sc

Mario A. Acero, A. Oliveros

Considering a well-motivated $f(R)$ modified-gravity model, in which an exponential function of the curvature is included, in this paper we implement a statistical data analysis to set constraints on the parameters of the model, taking into account an analytic approximate solution for the expansion rate, $H(z)$. Using a Monte Carlo Markov Chain-based analysi

Andrey Smirnov

We consider a class of $q$-hypergeometric equations describing the quantum difference equation for the cotangent bundles over projective spaces $X=T^{*}\mathbb{P}^{n-1}$ . We show that over $\mathbb{Q}_p$ these equations are equipped with the Frobenius action $(q,z)\to (q^p,z^p)$. We obtain an explicit formula for the constant term of the Frobenius intertwin

Theo F. Motta, Julian Bernhardt, Michael Buballa, Christian S. Fischer

In this work we continue our efforts to study the existence of a phase with an inhomogeneous, i.e., spatially varying, chiral condensate in QCD. To this end we employ a previously established method of stability analysis of the two-particle irreducible effective action in a truncation that corresponds to a rainbow-ladder approximation of the quark-gluon inte

Adam E. A. Fouda, Stephen H. Southworth, Phay J. Ho

The fragmentation of molecular cations following inner-shell decay processes in molecules containing heavy elements underpins the x-ray damage effects observed in x-ray scattering measurements of biological and chemical materials, as well as in medical applications involving Auger-electron emitting radionuclides. Traditionally, these processes are modeled us

Blake Temple, Robin Young

We prove the existence of ``pure tone&#39;&#39; nonlinear sound waves of all frequencies. These are smooth, time periodic, oscillatory solutions of the $3\times3$ compressible Euler equations satisfying periodic or acoustic boundary conditions in one space dimension. This resolves a centuries old problem in the theory of Acoustics, by establishing that the p

Gregory Schwartzman

We report that ChatGPT 4 and 4o are susceptible to a prompt injection attack that allows an attacker to exfiltrate users&#39; personal data. It is applicable without the use of any 3rd party tools and all users are currently affected. This vulnerability is exacerbated by the recent introduction of ChatGPT&#39;s memory feature, which allows an attacker to com

Atul Mohan, Nat Gopalswamy, Anshu Kumari, Sachiko Akiyama

Decameter hectometric (DH; 1-14 MHz) type-IV radio bursts are produced by flare-accelerated electrons trapped in post-flare loops or the moving magnetic structures associated with the CMEs. From a space weather perspective, it is important to systematically compile these bursts, explore their spectro-temporal characteristics, and study the associated CMEs. W

Yanting Teng, Rhine Samajdar, Katherine Van Kirk, Frederik Wilde

Learning faithful representations of quantum states is crucial to fully characterizing the variety of many-body states created on quantum processors. While various tomographic methods such as classical shadow and MPS tomography have shown promise in characterizing a wide class of quantum states, they face unique limitations in detecting topologically ordered

Maressa P Sampaio, Renan G Alvim, Felipe K Kalil, Maria C O Aguiar

Motivated by a problem from the 2023 International Physicists&#39; Tournament, we investigate the formation of particular patterns when light passes through glass. Experimentally, we use various glass plates, registering each reflected and transmitted outcome. Thus, we find the condition to form a common pattern, namely, randomly scratched plates produce the

Ranjeet Kumar, Newton Nath, Rahul Srivastava

We introduce a framework for hybrid neutrino mass generation, wherein scotogenic dark sector particles, including dark matter, are charged non-trivially under the $A_4$ flavor symmetry. The spontaneous breaking of the $A_4$ group to residual $\mathcal{Z}_2$ subgroup results in the ``cutting&#34; of the radiative loop. As a consequence the neutrinos acquire m

Hang Yu, Fei Dai

WASP-107 b seems to be a poster child of the long-suspected high-eccentricity migration scenario. It is on a 5.7-day, polar orbit. The planet is Jupiter-like in radius but Neptune-like in mass with exceptionally low density. WASP-107 c is on a 1100-day, $e=0.28$ orbit with at least Saturn mass. Planet b may still have a residual eccentricity of $0.06\pm 0.04

Sebastien Röcken, Anton F. Burnet, Julija Zavadlav

Machine learning (ML) potentials are a powerful tool in molecular modeling, enabling ab initio accuracy for comparably small computational costs. Nevertheless, all-atom simulations employing best-performing graph neural network architectures are still too expensive for applications requiring extensive sampling, such as free energy computations. Implicit solv

Matthew Buican

Coulomb branches of vacua are the most universal moduli spaces that arise in local unitary interacting 4d $\mathcal{N}=2$ superconformal field theories (SCFTs). In these theories, $1/2$-BPS primaries parameterize the Coulomb branches and form (anti-)chiral rings. We define the notion of a Coulomb branch operator algebra, $\mathcal{A}_{\mathcal{C}}$, that con

Graciana Puentes

We present an analytical and numerical study of a class of geometric phase induced by weak measurements. In particular, we analyze the dependence of the geometric phase on the winding ($W$) of the polar angle ($φ$), upon a sequence of $N$ weak measurements of increased magnitude ($c$), resulting in the appearance of a multiplicity of critical measurement-str

Hong Qian, Zhongwei Shen

Entropy, its production, and its change in a dynamical system can be understood from either a fully stochastic dynamic description or from a deterministic dynamics exhibiting chaotic behavior. By taking the former approach based on the general diffusion process with diffusion $\tfrac{1}α{\bf D}(\bf x)$ and drift $\bf b(\bf x)$, where $α$ represents the ``siz

Slavek M. Rucinski

A re-examination of high-resolution spectral monitoring of the W UMa-type binaries AW UMa and Epsilon CrA casts doubt on the widely utilized Lucy (1968a, 1968b) model of contact binaries. The detection of the very faint profile of the secondary component in AW UMa leads to a new spectroscopic determination of the mass ratio, q(sp) = 0.092 +/- 0.007, which is

Benjamin Thérien, Charles-Étienne Joseph, Boris Knyazev, Edouard Oyallon

Learned optimizers (LOs) have the potential to significantly reduce the wall-clock training time of neural networks. However, they can struggle to optimize unseen tasks (meta-generalize), especially when training networks wider than those seen during meta-training. To address this, we derive the Maximal Update Parametrization ($μ$P) for two state-of-the-art

Eric Balkanski, Steven DiSilvio, Alan Kuhnle, ChunLi Peng

In this work, we study the classical problem of maximizing a submodular function subject to a matroid constraint. We develop deterministic algorithms that are very parsimonious with respect to querying the submodular function, for both the case when the submodular function is monotone and the general submodular case. In particular, we present a 1/4 approxima

Alireza Fallah, Michael I. Jordan, Annie Ulichney

We consider a dynamic mechanism design problem where an auctioneer sells an indivisible good to groups of buyers in every round, for a total of $T$ rounds. The auctioneer aims to maximize their discounted overall revenue while adhering to a fairness constraint that guarantees a minimum average allocation for each group. We begin by studying the static case (

Xiaolong Sun, Liushuai Shi, Le Wang, Sanping Zhou

Temporal sentence grounding is a challenging task that aims to localize the moment spans relevant to a language description. Although recent DETR-based models have achieved notable progress by leveraging multiple learnable moment queries, they suffer from overlapped and redundant proposals, leading to inaccurate predictions. We attribute this limitation to t

Sara Willhammar, Hiroki Iimori, Joao Vieira, Lars Sundström

In what ways could cellular massive MIMO be improved? This technology has already been shown to bring huge performance gains. However, coverage holes and difficulties to transmit multiple streams to multi-antenna users because of insufficient channel rank remain issues. Distributed MIMO, also known as cell-free massive MIMO, might be the ultimate solution. H

Thomas Creutzig, Niklas Garner, Heeyeon Kim

We introduce a family of 3d $\mathcal{N} = 4$ superconformal field theories that have zero-dimensional Coulomb and Higgs branches and propose that the rational vertex operator algebras $W^{\text{min}}_{k - \scriptstyle{\frac{1}{2}}}(\mathfrak{sp}_{2N})$ and $L_{k}(\mathfrak{osp}_{1|2N})$ model the modular tensor categories of line operators in their topologi

Zhuo Chen, Rumen Dangovski, Charlotte Loh, Owen Dugan

We propose Quantum-informed Tensor Adaptation (QuanTA), a novel, easy-to-implement, fine-tuning method with no inference overhead for large-scale pre-trained language models. By leveraging quantum-inspired methods derived from quantum circuit structures, QuanTA enables efficient high-rank fine-tuning, surpassing the limitations of Low-Rank Adaptation (LoRA)-

Kevin Christian Wibisono, Yixin Wang

Large language models (LLMs) like transformers demonstrate impressive in-context learning (ICL) capabilities, allowing them to make predictions for new tasks based on prompt exemplars without parameter updates. While existing ICL theories often assume structured training data resembling ICL tasks (e.g., x-y pairs for linear regression), LLMs are typically tr

Robert Graf, Paul-Sören Platzek, Evamaria Olga Riedel, Constanze Ramschütz

Objectives: To present a publicly available deep learning-based torso segmentation model that provides comprehensive voxel-wise coverage, including delineations that extend to the boundaries of anatomical compartments. Materials and Methods: We extracted preliminary segmentations from TotalSegmentator, spine, and body composition models for Magnetic Resonanc

Valentin Vallucci-Goy, Ugo Lebreuilly, Patrick Hennebelle

The coupling between the magnetic field and the gas during the collapsing phase of star-forming cores is strongly affected by the dust size distribution, which is expected to evolve. We aim to investigate the influence of key parameters on the evolution of the dust distribution as well as on the magnetic resistivities during the protostellar collapse. We per

Mingyuan Meng, Dagan Feng, Lei Bi, Jinman Kim

Deformable image registration is a fundamental step for medical image analysis. Recently, transformers have been used for registration and outperformed Convolutional Neural Networks (CNNs). Transformers can capture long-range dependence among image features, which have been shown beneficial for registration. However, due to the high computation/memory loads

Corey Simmerer, Marisa Morakis, Lei Tian, Lia Gomez-Perez

Significance: Access to diagnostic eye care could be expanded with high-throughput and easy-to-use tools. Phase mask-based imaging may improve the fundus camera by enabling computational refocusing with no moving parts. While phase mask-based imaging has been demonstrated in a model eye, this approach has not been shown $\textit{in vivo}$. Aim: A computation

Andrew C. Li, Zizhao Chen, Toryn Q. Klassen, Pashootan Vaezipoor

Reward Machines provide an automaton-inspired structure for specifying instructions, safety constraints, and other temporally extended reward-worthy behaviour. By exposing the underlying structure of a reward function, they enable the decomposition of an RL task, leading to impressive gains in sample efficiency. Although Reward Machines and similar formal sp

Dionysios Karagiannis, Roy Maartens, Shun Saito, José Fonseca

A major goal of cosmology is to understand the nature of the field(s) which drove primordial Inflation. Through future observations, the statistics of large-scale structure will allow us to probe primordial non-Gaussianity of the curvature perturbation at the end of Inflation. We show how a new correlation statistic can significantly improve these constraint

Eura Nofshin, Esther Brown, Brian Lim, Weiwei Pan

Explanations of an AI&#39;s function can assist human decision-makers, but the most useful explanation depends on the decision&#39;s context, referred to as the downstream task. User studies are necessary to determine the best explanations for each task. Unfortunately, testing every explanation and task combination is impractical, especially considering the

Bao Liu, Tianbao Liu, Zhongshuo Hu, Fei Ye

The advancement of industrialization has spurred the development of innovative swarm intelligence algorithms, with Lion Swarm Optimization (LSO) notable for its robustness, parallelism, simplicity, and efficiency. While LSO excels in single-objective optimization, its multi-objective variants face challenges such as poor initialization, local optima entrapme

Ciamac C. Moallemi, Dan Robinson

Milionis et al.(2023) studied the rate at which automated market makers leak value to arbitrageurs when block times are discrete and follow a Poisson process, and where the risky asset price follows a geometric Brownian motion. We extend their model to analyze another popular mechanism in decentralized finance for onchain trading: Dutch auctions. We compute

Shenming Fu, Thomas Matheson, Aaron Meisner, Yuanyuan Zhang

We introduce a pipeline that performs rapid image subtraction and source selection to detect transients, with a focus on identifying gravitational wave optical counterparts using the Dark Energy Camera (DECam). In this work, we present the pipeline steps from processing raw data to identification of astrophysical transients on individual exposures. We proces

Bin Chen, Xiangliang Kong, Sijie Yu, Chengcai Shen

Where and how flares efficiently accelerate charged particles remains an unresolved question. Recent studies revealed that a &#34;magnetic bottle&#34; structure, which forms near the bottom of a large-scale reconnection current sheet above the flare arcade, is an excellent candidate for confining and accelerating charged particles. However, further understan

Thayne Currie

Recently, Biddle et al. (2024) claimed a non-detection of the protoplanet AB Aurigae b in Keck/NIRC2 Pa$β$ imaging. I reprocess these newly-public data and compare them to data from the extreme AO platform (SCExAO/CHARIS) used to discover AB Aur b. AB Aur b is decisively imaged with SCExAO/CHARIS at wavelengths covering Pa$β$. The Biddle et al. non detection

Francisco M. Blanco

We consider a class of finite-dimensional dynamical systems whose equations of motion are derived from a non-local-in-time action principle. The action functional has a zeroth order piece derived from a local Hamiltonian and a perturbation in the form of a non-local functional of the trajectory on phase space. We prove that the dynamics of these systems admi

X. Zhang, E. Bulbul, N. Malavasi, V. Ghirardini

The properties of the warm-hot intergalactic medium (WHIM) in cosmic filaments are among the least quantified units in modern astrophysics. The Spectrum Roentgen Gamma/eROSITA All Sky Survey ((SRG/eRASS) provides a unique opportunity to study the X-ray emission of the WHIM. We applied both imaging and spectroscopic stacking techniques to the data of the firs

Samuel Duffield, Kaelan Donatella, Johnathan Chiu, Phoebe Klett

Although theoretically compelling, Bayesian learning with modern machine learning models is computationally challenging since it requires approximating a high dimensional posterior distribution. In this work, we (i) introduce posteriors, an easily extensible PyTorch library hosting general-purpose implementations making Bayesian learning accessible and scala

Shi-Fan Chen, Zvonimir Vlah, Martin White

We study the bispectrum in Lagrangian perturbation theory. Extending past results for the power spectrum, we describe a method to efficiently compute the bispectrum in LPT, focusing on the Zeldovich approximation, in which contributions due to linear displacements are captured to all orders in a manifestly infrared (IR) safe way. We then isolate the effects

Jiaxuan Li, Jenny E. Greene, Scott G. Carlsten, Shany Danieli

It is well-known that almost all isolated dwarf galaxies are actively forming stars. We report the discovery of dw1322m2053 (nicknamed Hedgehog), an isolated quiescent dwarf galaxy at a distance of $2.40\pm0.15$ Mpc with a stellar mass of $M_\star \approx 10^{5.8}\, M_\odot$. The distance is measured using surface brightness fluctuations with both Legacy Sur

Charlie Cresswell-Hogg, Daniel F. Litim

We study the generation of fermion mass in a context where interactions break a discrete chiral symmetry. Then, fermion mass is not protected by a symmetry, no symmetry is broken by the generation of mass, and a vanishing mass no longer enhances a symmetry. We elaborate these scenarios for template fermionic and Yukawa theories in three dimensions where mass

David Stefanyszyn, Xi Tong, Yuhang Zhu

Cosmological correlators encode invaluable information about the wavefunction of the primordial universe. In this letter we present a duality between correlators and wavefunction coefficients that is valid to all orders in the loop expansion and manifests itself as a $\mathbb{Z}_4$ symmetry. To demonstrate the power of the duality, we derive a correlator-to-

Keisuke Inomata, Marc Kamionkowski, Celia M. Toral, Stephen R. Taylor

We present an efficient technique for calculating the angular two-point correlation functions (or ``overlap reduction functions&#39;&#39;) induced by gravitational waves in both the pulse arrival times of pulsars and in the angular deflections of distant sources. In the most general case, there are six auto- and cross-correlations for the pulse arrival times

Zeyi Sun, Tong Wu, Pan Zhang, Yuhang Zang

Recent years have witnessed remarkable progress in multi-view diffusion models for 3D content creation. However, there remains a significant gap in image quality and prompt-following ability compared to 2D diffusion models. A critical bottleneck is the scarcity of high-quality 3D objects with detailed captions. To address this challenge, we propose Bootstrap

Chaoyou Fu, Yuhan Dai, Yongdong Luo, Lei Li

In the quest for artificial general intelligence, Multi-modal Large Language Models (MLLMs) have emerged as a focal point in recent advancements. However, the predominant focus remains on developing their capabilities in static image understanding. The potential of MLLMs in processing sequential visual data is still insufficiently explored, highlighting the

Xiao Zhang, William Gao, Seemandhar Jain, Michael Maire

Image relighting is the task of showing what a scene from a source image would look like if illuminated differently. Inverse graphics schemes recover an explicit representation of geometry and a set of chosen intrinsics, then relight with some form of renderer. However error control for inverse graphics is difficult, and inverse graphics methods can represen

Hrushikesh Sable, Nathan M. Myers, Vito W. Scarola

A topological quantum number, the Witten index, characterizes supersymmetric models by probing for zero energy modes and the possibility of supersymmetry breaking. We propose an averaging method to infer the Witten index in quantum analog simulators. Motivated by recent work on Rydberg atoms trapped in optical tweezer arrays, we consider a related supersymme

Philip Lynch, Vojtěch Witzany, Maarten van de Meent, Niels Warburton

Extreme mass ratio inspirals (EMRIs), where a compact object orbits a massive black hole, are a key source of gravitational waves for the future Laser Interferometer Space Antenna (LISA). Due to their small mass ratio, ($ε\sim 10^{-4}$--$10^{-7}$), the binary evolves slowly and EMRI signals will be in-band for years. Additionally, astrophysical EMRIs are exp

Xin Wen, Bingchen Zhao, Yilun Chen, Jiangmiao Pang

Severe data imbalance naturally exists among web-scale vision-language datasets. Despite this, we find CLIP pre-trained thereupon exhibits notable robustness to the data imbalance compared to supervised learning, and demonstrates significant effectiveness in learning generalizable representations. With an aim to investigate the reasons behind this finding, w

Jean-Marc Valin, Ahmed Mustafa, Jan Büthe

Neural vocoders are now being used in a wide range of speech processing applications. In many of those applications, the vocoder can be the most complex component, so finding lower complexity algorithms can lead to significant practical benefits. In this work, we propose FARGAN, an autoregressive vocoder that takes advantage of long-term pitch prediction to

Kazimierz J. Borkowski, Stephen P. Reynolds, Robert Petre, David A. Green

We report results from 13 years of Chandra monitoring of nonthermal X-ray emission from the youngest Galactic supernova remnant G1.9+0.3, the only remnant known to be increasing in brightness. We confirm the spatially-integrated flux increase rate of $(1.2 \pm 0.2)$% yr$^{-1}$ between 1 and 7 keV, but find large spatial variations, from decreases of $-3$% yr

Hannah Lange, Guillaume Bornet, Gabriel Emperauger, Cheng Chen

Owing to their great expressivity and versatility, neural networks have gained attention for simulating large two-dimensional quantum many-body systems. However, their expressivity comes with the cost of a challenging optimization due to the in general rugged and complicated loss landscape. Here, we present a hybrid optimization scheme for neural quantum sta

Siyi Hu, Diego Martin Arroyo, Stephanie Debats, Fabian Manhardt

Generating realistic 3D scenes is an area of growing interest in computer vision and robotics. However, creating high-quality, diverse synthetic 3D content often requires expert intervention, making it costly and complex. Recently, efforts to automate this process with learning techniques, particularly diffusion models, have shown significant improvements in

Alan Bregazzi, James P. McGilligan, Paul F. Griffin, Erling Riis

Ultracold atoms are crucial for unlocking truly precise and accurate quantum metrology, and provide an essential platform for quantum computing, communication and memories. One of the largest ongoing challenges is the miniaturization of these quantum devices. Here, we show that the typically macroscopic optical lattice architecture at the heart of many ultra

Nicolas Zucchet, Antonio Orvieto

Recurrent neural networks (RNNs) notoriously struggle to learn long-term memories, primarily due to vanishing and exploding gradients. The recent success of state-space models (SSMs), a subclass of RNNs, to overcome such difficulties challenges our theoretical understanding. In this paper, we delve into the optimization challenges of RNNs and discover that,

Zhouxing Shi, Qirui Jin, Zico Kolter, Suman Jana

Branch-and-bound (BaB) is among the most effective techniques for neural network (NN) verification. However, existing works on BaB for NN verification have mostly focused on NNs with piecewise linear activations, especially ReLU networks. In this paper, we develop a general framework, named GenBaB, to conduct BaB on general nonlinearities to verify NNs with

Jianqing Liang, Min Chen, Jiye Liang

The Transformer architecture has recently gained considerable attention in the field of graph representation learning, as it naturally overcomes several limitations of Graph Neural Networks (GNNs) with customized attention mechanisms or positional and structural encodings. Despite making some progress, existing works tend to overlook external information of

Matthias Rosenkranz, Eric Brunner, Gabriel Marin-Sanchez, Nathan Fitzpatrick

A fundamental step of any quantum algorithm is the preparation of qubit registers in a suitable initial state. Often qubit registers represent a discretization of continuous variables and the initial state is defined by a multivariate function. We develop protocols for preparing quantum states whose amplitudes encode multivariate functions by linearly combin

Christopher Eckner, Felipe Figueroa, Piotr Tourkine

Regge poles connect the analytic structure of scattering amplitudes, analytically continued in spin, to the high-energy limit in momentum space. Dual models are expected to have only Regge poles, and string theory suggests there should be an infinite number of them. In this study, we investigate the number of Regge trajectories these models may have. We prov

Annette Huber, Martin Kalck

We apply the structure theory of finite dimensional algebras in order to deduce dimension formulas for spaces of period numbers, i.e., complex numbers defined by integrals of algebraic nature. We get a complete and conceptually clear answer in the case of $1$-periods, generalising classical results like Baker&#39;s theorem on the logarithms of algebraic numb

Carsten Lemmen, Philipp Sebastian Sommer

Frequently in socio-environmental sciences, models are used as tools to represent, understand, project and predict the behaviour of these complex systems. Along the modelling chain, Good Modelling Practices have been evolving that ensure - amongst others - that models are transparent and their results replicable. Whenever such models are represented in softw

Jim Fuller, Daichi Tsuna

The mass loss mechanism of red supergiant stars is not well understood, even though it has crucial consequences for their stellar evolution and the appearance of supernovae that occur upon core-collapse. We argue that outgoing shock waves launched near the photosphere can support a dense chromosphere between the star&#39;s surface and the dust formation radi

Kanghee Park, Jiayu Wang, Taylor Berg-Kirkpatrick, Nadia Polikarpova

Large Language Models (LLMs) struggle with reliably generating highly structured outputs, such as program code, mathematical formulas, or well-formed markup. Constrained decoding approaches mitigate this problem by greedily restricting what tokens an LLM can output at each step to guarantee that the output matches a given constraint. Specifically, in grammar

Fengdi Che, Chenjun Xiao, Jincheng Mei, Bo Dai

We prove that the combination of a target network and over-parameterized linear function approximation establishes a weaker convergence condition for bootstrapped value estimation in certain cases, even with off-policy data. Our condition is naturally satisfied for expected updates over the entire state-action space or learning with a batch of complete traje

Kieran A. Murphy, Sam Dillavou, Dani S. Bassett

Probabilistic representation spaces convey information about a dataset and are shaped by factors such as the training data, network architecture, and loss function. Comparing the information content of such spaces is crucial for understanding the learning process, yet most existing methods assume point-based representations, neglecting the distributional nat

Santiago Hernández-Gómez, Takuya Isogawa, Alessio Belenchia, Amikam Levy

The Kirkwood-Dirac quasiprobability distribution, intimately connected with the quantum correlation function of two observables measured at distinct times, is becoming increasingly relevant for fundamental physics and quantum technologies. This quasiprobability distribution can take non-positive values, and its experimental reconstruction becomes challenging

Fabian Obster, Christian Heumann

This paper introduces a novel framework for reducing variable selection bias by balancing selection frequencies of base-learners in boosting and introduces the sgboost package in R, which implements this framework combined with sparse-group boosting. The group bias reduction algorithm employs a simulation-based approach to iteratively adjust the degrees of f

Praveen Mishra, Arun Singh Patel, Sanjay Kumar Chauhan, Anirban Chakraborti

In this paper, we have explored the charge transfer mechanism in transition metal dichalcogenide (TMDC) quantum dots (QDs) of molybdenum disulfide ($\rm{MoS_2}$) and tungsten disulfide ($\rm{WS_2}$). Rhodamine 6G (R6G), a dye from the rhodamine family, has been employed as the fluorescent molecule, with MoS$_2$ and WS$_2$ QDs acting as electron acceptors in

Joseph S. B. Mitchell, Linh Nguyen

The well-known \textsc{Watchman Route} problem seeks a shortest route in a polygonal domain from which every point of the domain can be seen. In this paper, we study the cooperative variant of the problem, namely the \textsc{$k$-Watchmen Routes} problem, in a simple polygon $P$. We look at both the version in which the $k$ watchmen must collectively see all

Guilherme Franzmann

The identification of physical subsystems in quantum mechanics as compared to classical mechanics poses significant conceptual challenges, especially in the context of quantum gravity. Traditional approaches associate quantum systems with classical ones localized in spacetime, using either Hilbert space factors for finite-dimensional systems or local operato

Daniel A. Herrmann, Benjamin A. Levinstein

As large language models (LLMs) continue to demonstrate remarkable abilities across various domains, computer scientists are developing methods to understand their cognitive processes, particularly concerning how (and if) LLMs internally represent their beliefs about the world. However, this field currently lacks a unified theoretical foundation to underpin

Marta Vicentini, Ettore Bernardi, Matteo Bordin, Ekaterina Moreva

Along the path of unifying quantum theory and general relativity, it has recently been argued that the observation of gravity-induced entanglement could demonstrate the quantum nature of gravity; a few experimental proposals have been advanced, but extreme technological requirements make their implementation still far. Here we propose a table-top interferome

Elias Stengel-Eskin, Peter Hase, Mohit Bansal

When answering questions, LLMs can convey not only an answer, but a level of confidence about the answer being correct. This includes explicit confidence markers (e.g. giving a numeric score) as well as implicit markers, like an authoritative tone or elaborating with additional knowledge. For LLMs to be trustworthy knowledge sources, the confidence they conv

Jiesong Lian, Yucong Huang, Chengdong Ma, Mingzhi Wang

For solving zero-sum games involving non-transitivity, a useful approach is to maintain a policy population to approximate the Nash Equilibrium (NE). Previous studies have shown that the Policy Space Response Oracles (PSRO) algorithm is an effective framework for solving such games. However, current methods initialize a new policy from scratch or inherit a s

Rüdiger Valk, Daniel Moldt

Cycloids are particular Petri nets for modelling processes of actions and events, belonging to the fundaments of Petri&#39;s general systems theory. Defined by four parameters they provide an algebraic formalism to describe strongly synchronized sequential processes. To further investigate their structure, reduction systems of cycloids are defined in the sty

Daniel Duffy, Itay Griniasty, John Biggins, Cyrus Mostajeran

Shape-programmed sheets morph from one surface into another upon activation by stimuli such as illumination, and have attracted much interest for their potential engineering applications, especially in soft robotics. Complex shape changes can be achieved by patterning a simple local active deformation (e.g. isotropic swelling), to generate differential growt

Dongho Shin, Yongyun Shin, Nao Hagiwara

We consider Bayesian estimation of a hierarchical linear model (HLM) from partially observed data, assumed to be missing at random, and small sample sizes. A vector of continuous covariates $C$ includes cluster-level partially observed covariates with interaction effects. Due to small sample sizes from 37 patient-physician encounters repeatedly measured at f

Xiaojun Jia, Tianyu Pang, Chao Du, Yihao Huang

Large language models (LLMs) are being rapidly developed, and a key component of their widespread deployment is their safety-related alignment. Many red-teaming efforts aim to jailbreak LLMs, where among these efforts, the Greedy Coordinate Gradient (GCG) attack&#39;s success has led to a growing interest in the study of optimization-based jailbreaking techn

Per Berglund, Giorgi Butbaia, Yang-Hui He, Elli Heyes

We apply reinforcement learning (RL) to generate fine regular star triangulations of reflexive polytopes, that give rise to smooth Calabi-Yau (CY) hypersurfaces. We demonstrate that, by simple modifications to the data encoding and reward function, one can search for CYs that satisfy a set of desirable string compactification conditions. For instance, we sho

Fatema Jannat Dihan, Saydul Akbar Murad

Monkeypox virus (MPXV) is a zoonotic virus that poses a significant threat to public health, particularly in remote parts of Central and West Africa. Early detection of monkeypox lesions is crucial for effective treatment. However, due to its similarity with other skin diseases, monkeypox lesion detection is a challenging task. To detect monkeypox, many rese

Ben Cottier, Robi Rahman, Loredana Fattorini, Nestor Maslej

The costs of training frontier AI models have grown dramatically in recent years, but there is limited public data on the magnitude and growth of these expenses. This paper develops a detailed cost model to address this gap, estimating training costs using three approaches that account for hardware, energy, cloud rental, and staff expenses. The analysis reve

Pengyuan Lyu, Yulin Li, Hao Zhou, Weihong Ma

Text-rich images have significant and extensive value, deeply integrated into various aspects of human life. Notably, both visual cues and linguistic symbols in text-rich images play crucial roles in information transmission but are accompanied by diverse challenges. Therefore, the efficient and effective understanding of text-rich images is a crucial litmus

Konstantin Hess, Dennis Frauen, Valentyn Melnychuk, Stefan Feuerriegel

Estimating potential outcomes for treatments over time based on observational data is important for personalized decision-making in medicine. However, many existing methods for this task fail to properly adjust for time-varying confounding and thus yield biased estimates. There are only a few neural methods with proper adjustments, but these have inherent li