May 2024 arXiv papers — page 10

Manish Tamta, Nirmal Raj, Prateek Sharma

Primordial black holes (PBHs) in the mass range $10^{-16}-10^{-11}~M_\odot$ may constitute all the dark matter. We show that gravitational microlensing of bright x-ray pulsars provide the most robust and immediately implementable opportunity to uncover PBH dark matter in this mass window. As proofs of concept, we show that the currently operational NICER tel

Kailu Wu, Fangfu Liu, Zhihan Cai, Runjie Yan

In this work, we introduce Unique3D, a novel image-to-3D framework for efficiently generating high-quality 3D meshes from single-view images, featuring state-of-the-art generation fidelity and strong generalizability. Previous methods based on Score Distillation Sampling (SDS) can produce diversified 3D results by distilling 3D knowledge from large 2D diffus

Muhammad Zubair Irshad

Recent advances in deep learning have led to a data-centric intelligence i.e. artificially intelligent models unlocking the potential to ingest a large amount of data and be really good at performing digital tasks such as text-to-image generation, machine-human conversation, and image recognition. This thesis covers the topic of learning with structured indu

Rachel D. Gidaro, Jane L. Harvill

Although many time series are realizations from discrete processes, it is often that a continuous Gaussian model is implemented for modeling and forecasting the data, resulting in incoherent forecasts. Forecasts using a Poisson-Lindley integer autoregressive (PLINAR) model are compared to variations of Gaussian forecasts via simulation by equating relevant m

Tal Reiss, George Kour, Naama Zwerdling, Ateret Anaby-Tavor

When first deploying an anomaly detection system, e.g., to detect out-of-scope queries in chatbots, there are no observed data, making data-driven approaches ineffective. Zero-shot anomaly detection methods offer a solution to such "cold-start" cases, but unfortunately they are often not accurate enough. This paper studies the realistic but underexplored col

Ling-Hao Chen, Shunlin Lu, Ailing Zeng, Hao Zhang

This study delves into the realm of multi-modality (i.e., video and motion modalities) human behavior understanding by leveraging the powerful capabilities of Large Language Models (LLMs). Diverging from recent LLMs designed for video-only or motion-only understanding, we argue that understanding human behavior necessitates joint modeling from both videos an

Feipeng Ma, Hongwei Xue, Guangting Wang, Yizhou Zhou

Existing Multimodal Large Language Models (MLLMs) follow the paradigm that perceives visual information by aligning visual features with the input space of Large Language Models (LLMs), and concatenating visual tokens with text tokens to form a unified sequence input for LLMs. These methods demonstrate promising results on various vision-language tasks but a

Paolo Piersanti, Tianyu Sun

This paper is devoted to the study of a novel mixed Finite Element Method for approximating the solutions of fourth order variational problems subjected to a constraint. The first problem we consider consists in establishing the convergence of the error of the numerical approximation of the solution of a biharmonic obstacle problem. The contents of this sect

Lening Wang, Wenzhao Zheng, Yilong Ren, Han Jiang

Understanding the evolution of 3D scenes is important for effective autonomous driving. While conventional methods mode scene development with the motion of individual instances, world models emerge as a generative framework to describe the general scene dynamics. However, most existing methods adopt an autoregressive framework to perform next-token predicti

Jiaben Chen, Xin Yan, Yihang Chen, Siyuan Cen

In this work, we introduce a challenging task for simultaneously generating 3D holistic body motions and singing vocals directly from textual lyrics inputs, advancing beyond existing works that typically address these two modalities in isolation. To facilitate this, we first collect the RapVerse dataset, a large dataset containing synchronous rapping vocals,

Bolin Ni, JingCheng Hu, Yixuan Wei, Houwen Peng

In this work, we present Xwin-LM, a comprehensive suite of alignment methodologies for large language models (LLMs). This suite encompasses several key techniques, including supervised finetuning (SFT), reward modeling (RM), rejection sampling finetuning (RS), and direct preference optimization (DPO). The key components are as follows: (1) Xwin-LM-SFT, model

Yao-Chih Lee, Yi-Ting Chen, Andrew Wang, Ting-Hsuan Liao

We introduce VividDream, a method for generating explorable 4D scenes with ambient dynamics from a single input image or text prompt. VividDream first expands an input image into a static 3D point cloud through iterative inpainting and geometry merging. An ensemble of animated videos is then generated using video diffusion models with quality refinement tech

Chinedu Innocent Nwoye, Nicolas Padoy

Accurate tool tracking is essential for the success of computer-assisted intervention. Previous efforts often modeled tool trajectories rigidly, overlooking the dynamic nature of surgical procedures, especially tracking scenarios like out-of-body and out-of-camera views. Addressing this limitation, the new CholecTrack20 dataset provides detailed labels that

Muldrow Etheredge, Ben Heidenreich, Tom Rudelius, Ignacio Ruiz

The Emergent String Conjecture constrains the possible types of light towers in infinite-distance limits in quantum gravity moduli spaces. In this paper, we use these constraints to restrict the geometry of the scalar charge-to-mass vectors $(-\vec{\nabla}\log m)$ of the light towers and the analogous vector $(-\vec{\nabla}\log\Lambda_{\text{QG}})$ of the sp

Laura Kopf, Philine Lou Bommer, Anna Hedström, Sebastian Lapuschkin

A crucial aspect of understanding the complex nature of Deep Neural Networks (DNNs) is the ability to explain learned concepts within their latent representations. While methods exist to connect neurons to human-understandable textual descriptions, evaluating the quality of these explanations is challenging due to the lack of a unified quantitative approach.

Dixuan Lin, Yuxiang Zhang, Mengcheng Li, Wei Jing

In this paper, we introduce OmniHands, a universal approach to recovering interactive hand meshes and their relative movement from monocular or multi-view inputs. Our approach addresses two major limitations of previous methods: lacking a unified solution for handling various hand image inputs and neglecting the positional relationship of two hands within im

Marcus DuPont, Andrew MacFadyen

We consider the observational implications of the binary neutron star (BNS) merger GW170817 leaving behind a rapidly rotating massive neutron star that launches a relativistic, equatorial outflow as well as a jet. We show that if the equatorial outflow (ring) is highly beamed in the equatorial plane, its luminosity can be "hidden" from view until late times,

Dimitris Alevras, Mihir Metkar, Takahiro Yamamoto, Vaibhaw Kumar

Recent advancements in quantum computing have opened new avenues for tackling long-standing complex combinatorial optimization problems that are intractable for classical computers. Predicting secondary structure of mRNA is one such notoriously difficult problem that can benefit from the ever-increasing maturity of quantum computing technology. Accurate pred

Chen Wang, Jiatao Gu, Xiaoxiao Long, Yuan Liu

Recent years have seen significant advancements in 3D generation. While methods like score distillation achieve impressive results, they often require extensive per-scene optimization, which limits their time efficiency. On the other hand, reconstruction-based approaches are more efficient but tend to compromise quality due to their limited ability to handle

Anupam A. H, P. V. Athira, Priyadarshi Paul, Suvrat Raju

We study the properties of massive fields extrapolated to the blowup of spatial infinity ($\hat{i}^0$), extending the program initiated in arXiv:2207.06406. In the free theory, we find an explicit representation of boundary two-point functions and boundary to bulk two-point functions, and also present an HKLL-type reconstruction formula for local bulk operat

Shuyuan Tu, Qi Dai, Zihao Zhang, Sicheng Xie

Despite impressive advancements in diffusion-based video editing models in altering video attributes, there has been limited exploration into modifying motion information while preserving the original protagonist's appearance and background. In this paper, we propose MotionFollower, a lightweight score-guided diffusion model for video motion editing. To intr

Zhiqiang Wang, Dejia Xu, Rana Muhammad Shahroz Khan, Yanbin Lin

Image geolocation is a critical task in various image-understanding applications. However, existing methods often fail when analyzing challenging, in-the-wild images. Inspired by the exceptional background knowledge of multimodal language models, we systematically evaluate their geolocation capabilities using a novel image dataset and a comprehensive evaluat

Nicolas Dufour, Victor Besnier, Vicky Kalogeiton, David Picard

Conditional diffusion models are powerful generative models that can leverage various types of conditional information, such as class labels, segmentation masks, or text captions. However, in many real-world scenarios, conditional information may be noisy or unreliable due to human annotation errors or weak alignment. In this paper, we propose the Coherence-

Nan Huang, Xiaobao Wei, Wenzhao Zheng, Pengju An

Photorealistic 3D reconstruction of street scenes is a critical technique for developing real-world simulators for autonomous driving. Despite the efficacy of Neural Radiance Fields (NeRF) for driving scenes, 3D Gaussian Splatting (3DGS) emerges as a promising direction due to its faster speed and more explicit representation. However, most existing street 3

András Gilyén, Chi-Fang Chen, Joao F. Doriguello, Michael J. Kastoryano

Classical Markov Chain Monte Carlo methods have been essential for simulating statistical physical systems and have proven well applicable to other systems with many degrees of freedom. Motivated by the statistical physics origins, Chen, Kastoryano, and Gily\'en [CKG23] proposed a continuous-time quantum thermodynamic analogue to Glauber dynamics that is (i)

Yifeng Zhu, Arisrei Lim, Peter Stone, Yuke Zhu

This work presents an object-centric approach to learning vision-based manipulation skills from human videos. We investigate the problem of robot manipulation via imitation in the open-world setting, where a robot learns to manipulate novel objects from a single video demonstration. We introduce ORION, an algorithm that tackles the problem by extracting an o

Varun Magesh, Faiz Surani, Matthew Dahl, Mirac Suzgun

Legal practice has witnessed a sharp rise in products incorporating artificial intelligence (AI). Such tools are designed to assist with a wide range of core legal tasks, from search and summarization of caselaw to document drafting. But the large language models used in these tools are prone to "hallucinate," or make up false information, making their use r

Sangyun Lee, Zinan Lin, Giulia Fanti

Diffusion models have shown great promise for image and video generation, but sampling from state-of-the-art models requires expensive numerical integration of a generative ODE. One approach for tackling this problem is rectified flows, which iteratively learn smooth ODE paths that are less susceptible to truncation error. However, rectified flows still requ

Aditya Ganeshan, Ryan Y. Huang, Xianghao Xu, R. Kenny Jones

The ability to edit 3D assets from natural language presents a compelling paradigm to aid in the democratization of 3D content creation. However, while natural language is often effective at communicating general intent, it is poorly suited for specifying precise manipulation. To address this gap, we introduce ParSEL, a system that enables controllable editi

Roberto Ceraolo, Dmitrii Kharlapenko, Ahmad Khan, Amélie Reymond

Recent progress in Large Language Model (LLM) technology has changed our role in interacting with these models. Instead of primarily testing these models with questions we already know answers to, we are now using them for queries where the answers are unknown to us, driven by human curiosity. This shift highlights the growing need to understand curiosity-dr

Subhankar Mahapatra, Santanu Sarkar

This paper discusses an abstract Kramer sampling theorem for functions within a reproducing kernel Hilbert space (RKHS) of vector valued holomorphic functions. Additionally, we extend the concept of quasi Lagrange-type interpolation for functions within a RKHS of vector valued entire functions. The dependence of having quasi Lagrange-type interpolation on an

Samantha L. Garza, Jessica K. Werk, Benjamin D. Oppenheimer, Kirill Tchernyshyov

We present an analysis of \textit{HST}/COS/G160M observations of CIV in the inner circumgalactic medium (CGM) of a novel sample of eight z$\sim$0, L$\approx$L$^{\star}$ galaxies, paired with UV-bright QSOs at impact parameters ($R_\mathrm{proj}$) between 25-130 kpc. The galaxies in this stellar-mass-controlled sample (log$_{10}$M$_{\star}$/M$_{\odot}$ $\sim$

Ziwei Ji, Yuzhe Gu, Wenwei Zhang, Chengqi Lyu

Reducing the `$\textit{hallucination}$' problem of Large Language Models (LLMs) is crucial for their wide applications. A comprehensive and fine-grained measurement of the hallucination is the first key step for the governance of this issue but is under-explored in the community. Thus, we present $\textbf{ANAH}$, a bilingual dataset that offers $\textbf{AN}$

Wei Zhong, Manasa Bharadwaj

Speculative decoding (SD) has attracted a significant amount of research attention due to the substantial speedup it can achieve for LLM inference. However, despite the high speedups they offer, speculative decoding methods often achieve optimal performance on high-end devices or with a substantial GPU memory overhead. Given limited memory and the necessity

Guillaume Huguet, James Vuckovic, Kilian Fatras, Eric Thibodeau-Laufer

Proteins are essential for almost all biological processes and derive their diverse functions from complex 3D structures, which are in turn determined by their amino acid sequences. In this paper, we exploit the rich biological inductive bias of amino acid sequences and introduce FoldFlow-2, a novel sequence-conditioned SE(3)-equivariant flow matching model

Erik J. Wasleske, Vivienne F. Baldassare

We assemble a sample of 733 dwarf galaxies ($M_{\ast} \le 10^{9.5} \text{M}_\odot$) with signatures of active galactic nuclei (AGN) and explore the intersection between different AGN selection techniques. Objects in our database are compiled from previous studies that identify AGN in dwarf galaxies through spectroscopy, X-ray emission, infrared colors, and o

Ofir Gorodetsky, Mo Dick Wong

Let $\alpha$ be a Steinhaus or a Rademacher random multiplicative function. For a wide class of multiplicative functions $f$ we show that the sum $\sum_{n \le x}\alpha(n) f(n)$, normalised to have mean square $1$, has a non-Gaussian limiting distribution. More precisely, we establish a generalised central limit theorem with random variance determined by the

Jianghao Shen, Nan Xue, Tianfu Wu

Learning 3D scene representation from a single-view image is a long-standing fundamental problem in computer vision, with the inherent ambiguity in predicting contents unseen from the input view. Built on the recently proposed 3D Gaussian Splatting (3DGS), the Splatter Image method has made promising progress on fast single-image novel view synthesis via lea

Ajay Patel, Markus Hofmarcher, Claudiu Leoveanu-Condrei, Marius-Constantin Dinu

Training models to act as agents that can effectively navigate and perform actions in a complex environment, such as a web browser, has typically been challenging due to lack of training data. Large language models (LLMs) have recently demonstrated some capability to navigate novel environments as agents in a zero-shot or few-shot fashion, purely guided by n

Ashwin Sah, Julian Sahasrabudhe, Mehtaab Sawhney

Let $M$ be an $n\times n$ matrix with iid subgaussian entries with mean $0$ and variance $1$ and let $\sigma_n(M)$ denote the least singular value of $M$. We prove that \[\mathbb{P}\big( \sigma_{n}(M) \leq \varepsilon n^{-1/2} \big) = (1+o(1)) \varepsilon + e^{-\Omega(n)}\] for all $0 \leq \varepsilon \ll 1$. This resolves, up to a $1+o(1)$ factor, a seminal

Bohao Li, Fengcheng Wu

We present a theoretical study of mapping between Chern bands and generalized Landau levels in twisted bilayer MoTe$_2$ ($t$MoTe$_2$), where fractional Chern insulators down to zero magnetic fields have been observed. We construct an exact Landau-level representation of moir\'e bands, where the basis functions, characterized by a uniform quantum geometry, ar

Levon Pogosian, Gong-Bo Zhao, Karsten Jedamzik

The value of the Hubble constant determined from CMB and BAO measurements is directly dependent on the sound horizon at the photon-baryon decoupling. There has been significant interest in the possibility of new physics at the epoch around recombination that could reduce the sound horizon and increase the inferred value of $H_0$, thus helping to relieve the

Himangi Mittal, Nakul Agarwal, Shao-Yuan Lo, Kwonjoon Lee

We introduce PlausiVL, a large video-language model for anticipating action sequences that are plausible in the real-world. While significant efforts have been made towards anticipating future actions, prior approaches do not take into account the aspect of plausibility in an action sequence. To address this limitation, we explore the generative capability o

Shyam Sundhar Ramesh, Yifan Hu, Iason Chaimalas, Viraj Mehta

Adapting large language models (LLMs) for specific tasks usually involves fine-tuning through reinforcement learning with human feedback (RLHF) on preference data. While these data often come from diverse labelers' groups (e.g., different demographics, ethnicities, company teams, etc.), traditional RLHF approaches adopt a "one-size-fits-all" approach, i.e.,

Mark de Berg, Rupert Freeman, Ulrike Schmidt-Kraepelin, Markus Utke

We study a budget-aggregation setting in which a number of voters report their ideal distribution of a budget over a set of alternatives, and a mechanism aggregates these reports into an allocation. Ideally, such mechanisms are truthful, i.e., voters should not be incentivized to misreport their preferences. For the case of two alternatives, the set of mecha

Gan Zhang

Wind energy plays a critical role in mitigating climate change and meeting growing energy demands. However, the long-term impacts of anthropogenic warming on wind resources, particularly their seasonal variations and potential compounding risks, remain understudied. Here we analyze large-ensemble climate simulations in high-emission scenarios to assess the p

Daniel Pomerleano, Constantin Teleman

For a compact monotone symplectic manifold $X$ with Hamiltonian action of a compact Lie group $G$ and smooth symplectic reduction, we relate its gauged $2$-dimensional $A$-model to the $A$-model of $X/\!/G$. This (long conjectured) result is parallel to the ($B$-model!) \emph{quantization commutes with reduction} theorem of Guillemin and Sternberg in quantum

Karel Devriendt

Kemeny's constant is an invariant of discrete-time Markov chains, equal to the expected number of steps between two states sampled from the stationary distribution. It appears in applications as a concise characterization of the mixing properties of a Markov chain and has many alternative definitions. In this short article, we derive a new geometric expressi

Stefano Disca, Roberto Demaria, Lorenzo Formaggio

We present a collection of 12 exercises picked from the exam tests of the course "Elements of Quantum Field Theory", teached by professor Mauro Moretti in the academic year 2021-22 for the Master's Degree in Physics at the University of Ferrara. The goal of this review is to provide a general method to approach the main types of calculations in Quantum Field

Jinrui Yang, Xianhang Li, Druv Pai, Yuyin Zhou

CRATE, a white-box transformer architecture designed to learn compressed and sparse representations, offers an intriguing alternative to standard vision transformers (ViTs) due to its inherent mathematical interpretability. Despite extensive investigations into the scaling behaviors of language and vision transformers, the scalability of CRATE remains an ope

Arist Zhenyuan Yang, Zheng-Xin Liu

Gapless quasiparticles can exist in the Bogoliubov-de Gennes (BdG) Hamiltonians in the mean field description of superconductors (SCs), fermionic superfluids (SFs) and quantum spin liquids (QSLs). The mechanism of gapless quasiparticles in superconductors was studied in literature based on the homotopy theory or symmetry-indicators. However, important proper

Valery V. Ryzhikov

There is a set of continual cardinality of pairwise disjoint Gaussian automorphisms with spectrally isomorphic even factors having Lebesgue spectrum.

Marco Adani, Simone Cavazzoni, Berihu Teklu, Paolo Bordone

We address quantum metrology in critical spin chains with anisotropy and Dzyaloshinskii-Moriya (DM) interaction, and show how local and quasi-local measurements may be exploited to characterize global properties of the systems. In particular, we evaluate the classical (magnetization) and quantum Fisher information of the relevant parameters for the density m

Longcheng Li, Qian Li, Xingjian Li, Qipeng Liu

The seminal work by Impagliazzo and Rudich (STOC'89) demonstrated the impossibility of constructing classical public key encryption (PKE) from one-way functions (OWF) in a black-box manner. However, the question remains: can quantum PKE (QPKE) be constructed from quantumly secure OWF? A recent line of work has shown that it is indeed possible to build QPKE f

Qipin Chen, Shane Chern, Lin Jiu

Lattice geometries and random walks on them are of great interest for their applications in different fields such as physics, chemistry, and computer science. In this work, we focus on multi-headed lattices and study properties of the Green functions for these lattices such as the associated differential equations and the P\'olya numbers. In particular, we c

Christian Bernert, Ulrich Derenthal

We prove Manin's conjecture for split smooth quintic del Pezzo surfaces over arbitrary number fields with respect to fairly general anticanonical height functions. After passing to universal torsors, we first show that we may restrict the torsor variables to their typical sizes, and then we can solve the counting problem in the framework of o-minimal structu

Song He, Yu-tin Huang, Chia-Kai Kuo

In this letter, we consider a positive geometry conjectured to encode the loop integrand of four-point stress-energy correlators in planar $\mathcal{N}=4$ super Yang-Mills. Beginning with four lines in twistor space, we characterize a positive subspace to which an $\ell$-loop geometry is attached. The loop geometry then consists of $\ell$ lines in twistor sp

Weilin Lin, Li Liu, Shaokui Wei, Jianze Li

The security threat of backdoor attacks is a central concern for deep neural networks (DNNs). Recently, without poisoned data, unlearning models with clean data and then learning a pruning mask have contributed to backdoor defense. Additionally, vanilla fine-tuning with those clean data can help recover the lost clean accuracy. However, the behavior of clean

E. S. Andriyakhina, S. L. Khortsev, F. Evers

As is well known, magnetic impurities adsorbed on superconductors, e.g. of the s-wave type, can introduce a bound gap-state (Yu-Shiba-Rusinov resonance). We here investigate within a minimal model how the impurity moment arranges with respect to a weak homogeneous internal magnetic field employing a fully self-consistent mean-field treatment. Our investigati

Zachary Novack, Julian McAuley, Taylor Berg-Kirkpatrick, Nicholas Bryan

Controllable music generation methods are critical for human-centered AI-based music creation, but are currently limited by speed, quality, and control design trade-offs. Diffusion Inference-Time T-optimization (DITTO), in particular, offers state-of-the-art results, but is over 10x slower than real-time, limiting practical use. We propose Distilled Diffusio

Paul M. Voutier

We produce an explicit family of totally real cyclic quartic polynomials that are monogenic in many cases and, if the $abc$ conjecture holds, generate distinct monogenic quartic fields infinitely often. Additional families (also conjecturally generating infinitely many distinct fields) are provided in Section 4, including what appears to be an infinite colle

Maria Bånkestad, Olof Mogren, Aleksis Pirinen

This paper presents a novel approach for constructing graph neural networks equivariant to 2D rotations and translations and leveraging them as PDE surrogates on non-gridded domains. We show that aligning the representations with the principal axis allows us to sidestep many constraints while preserving SE(2) equivariance. By applying our model as a surrogat

David Cui, Arthur Mehta, Denis Rochette

Bell monogamy relations characterize the trade-offs in Bell inequality violations among pairs of players in multiplayer settings. In this work, we introduce a method for extending monogamy relations from a distinguished set of configurations to monogamy relations on all possible multiplayer settings. Applying this approach, we show that nonlocality in the CH

Panagiotis C. Theocharopoulos, Spiros V. Georgakopoulos, Sotiris K. Tasoulis, Vassilis P. Plagianakos

With the increasing use of Artificial Intelligence in Natural Language Processing, concerns have been raised regarding the detection of AI-generated text in various domains. This study aims to investigate this issue by proposing a methodology to accurately distinguish AI-generated and human-written book reviews. Our approach utilizes transfer learning, enabl

Cédric Boutillier, Béatrice de Tilière

We consider the dimer model on the Aztec diamond with Fock's weights, which is gauge equivalent to the model with any choice of positive weight function. We prove an explicit, compact formula for the inverse Kasteleyn matrix, thus extending numerous results in the case of periodic graphs. We also show an explicit product formula for the partition function; a

Minghao Guo, Bohan Wang, Kaiming He, Wojciech Matusik

We introduce TetSphere Splatting, a Lagrangian geometry representation designed for high-quality 3D shape modeling. TetSphere splatting leverages an underused yet powerful geometric primitive -- volumetric tetrahedral meshes. It represents 3D shapes by deforming a collection of tetrahedral spheres, with geometric regularizations and constraints that effectiv

Chaoyang Wang, Xiangtai Li, Lu Qi, Henghui Ding

Semantic segmentation and semantic image synthesis are two representative tasks in visual perception and generation. While existing methods consider them as two distinct tasks, we propose a unified framework (SemFlow) and model them as a pair of reverse problems. Specifically, motivated by rectified flow theory, we train an ordinary differential equation (OD

Fangqi Dong, Qipeng Liu, Kewen Wu

Cryptography often considers the strongest yet plausible attacks in the real world. Preprocessing (a.k.a. non-uniform attack) plays an important role in both theory and practice: an efficient online attacker can take advantage of advice prepared by a time-consuming preprocessing stage. Salting is a heuristic strategy to counter preprocessing attacks by feedi

Di-Lun Yang, Xiaojun Yao

We study the spin-dependent in-medium dynamics of quarkonia by using the potential nonrelativistic QCD (pNRQCD) and the open quantum system framework. We consider the pNRQCD Lagrangian valid up to the order $\frac{r}{M^0}=r$ and $\frac{r^0}{M}=\frac{1}{M}$ in the double power counting. By considering the Markovian condition and applying the Wigner transforma

Sijie Zhao, Yong Zhang, Xiaodong Cun, Shaoshu Yang

Spatio-temporal compression of videos, utilizing networks such as Variational Autoencoders (VAE), plays a crucial role in OpenAI's SORA and numerous other video generative models. For instance, many LLM-like video models learn the distribution of discrete tokens derived from 3D VAEs within the VQVAE framework, while most diffusion-based video models capture

Dániel Rácz, Mihály Petreczky, Bálint Daróczy

Many state-of-the-art models trained on long-range sequences, for example S4, S5 or LRU, are made of sequential blocks combining State-Space Models (SSMs) with neural networks. In this paper we provide a PAC bound that holds for these kind of architectures with \emph{stable} SSM blocks and does not depend on the length of the input sequence. Imposing stabili

Meng Qin, Chaorui Zhang, Yu Gao, Weixi Zhang

Community detection (CD) is a classic graph inference task that partitions nodes of a graph into densely connected groups. While many CD methods have been proposed with either impressive quality or efficiency, balancing the two aspects remains a challenge. This study explores the potential of deep graph learning to achieve a better trade-off between the qual

Noah Halberstam, Tom Hutchcroft

We prove a conjecture of Diaconis and Freedman (Ann. Probab. 1980) characterising the extreme points of the set of partially-exchangeable processes on a countable set. More concretely, we prove that the partially exchangeable sigma-algebra of any transient partially exchangeable process $X=(X_i)_{i\geq 0}$ (and hence any transient Markov chain) coincides up

Renata Kallosh

U-duality imposes strong constraints on the structure of divergences in supergravity. But Gaillard-Zumino Sp$(2n_v, \mathbb{R})$ duality in D=4 has more symmetries than U-duality. For example, in $N = 8$ the dimension of Sp(56) is 1596, whereas its U-duality subgroup $E_{7(7)}$ has a dimension of 133. In comparison, in D $> 4$ maximal dualities are U-dualiti

Siva Uday Sampreeth Chebolu, Franck Dernoncourt, Nedim Lipka, Thamar Solorio

Aspect-Based Sentiment Analysis (ABSA) has experienced tremendous expansion and diversity due to various shared tasks spanning several languages and fields and organized via SemEval workshops and Germeval. Nonetheless, a few shortcomings still need to be addressed, such as the lack of low-resource language evaluations and the emphasis on sentence-level analy

Alvin Gonzales, Rebekah Herrman, Colin Campbell, Igor Gaidai

Continuous-time quantum walks (CTQWs) on dynamic graphs, referred to as dynamic CTQWs, are a recently introduced universal model of computation that offers a new paradigm in which to envision quantum algorithms. In this work we develop an algorithm that converts single-edge and self-loop dynamic CTQWs to the gate model of computation. We use this mapping to

Martin Bertran, Shuai Tang, Michael Kearns, Jamie Morgenstern

Machine unlearning is motivated by desire for data autonomy: a person can request to have their data's influence removed from deployed models, and those models should be updated as if they were retrained without the person's data. We show that, counter-intuitively, these updates expose individuals to high-accuracy reconstruction attacks which allow the attac

Massimo Bini, Karsten Roth, Zeynep Akata, Anna Khoreva

Parameter-efficient finetuning (PEFT) has become ubiquitous to adapt foundation models to downstream task requirements while retaining their generalization ability. However, the amount of additionally introduced parameters and compute for successful adaptation and hyperparameter searches can explode quickly, especially when deployed at scale to serve numerou

Michelle M. Kelley, Joshua Quinton, Kamron Fazel, Nima Karimitari

The accuracy of density-functional theory (DFT) is determined by the quality of the approximate functionals, such as exchange-correlation in electronic DFT and the excess functional in the classical DFT formalism of fluids. The exact functional is highly nonlocal for both electrons and fluids, yet most approximate functionals are semi-local or nonlocal in a

Guy Mor-Lan, Effi Levi, Tamir Sheafer, Shaul R. Shenhav

We introduce IsraParlTweet, a new linked corpus of Hebrew-language parliamentary discussions from the Knesset (Israeli Parliament) between the years 1992-2023 and Twitter posts made by Members of the Knesset between the years 2008-2023, containing a total of 294.5 million Hebrew tokens. In addition to raw text, the corpus contains comprehensive metadata on s

Xin Huang

We introduce a new type of equivalence between blocks of finite group algebras called an almost isotypy. An almost isotypy restricts to a weak isotypy in Brou\'{e}'s original definition, and it is slightly weaker than Linckelmann's version. We show that a bimodule of two block algebras of finite groups - which has an endopermutation module as a source and wh

Ruochen Zhao, Wenxuan Zhang, Yew Ken Chia, Weiwen Xu

As LLMs continuously evolve, there is an urgent need for a reliable evaluation method that delivers trustworthy results promptly. Currently, static benchmarks suffer from inflexibility and unreliability, leading users to prefer human voting platforms like Chatbot Arena. However, human evaluations require significant manual effort. To address this, we propose

Indranil Mukherjee, P. K. Mohanty

We show that equilibrium systems in $d$ dimension that obey the inequality $d\nu> 2,$ known as Harris criterion, exhibit suppressed energy fluctuation in their critical state. Ashkin-Teller model is an example in $d=2$ where the correlation length exponent $\nu$ varies continuously with the inter-spin interaction strength $\lambda$ and exceeds the value $\fr

Arnd Scheel, Qiliang Wu

We investigate the effect of spatial inhomogeneity on perfectly periodic, self-organized striped patterns in spatially extended systems. We demonstrate that inhomogeneities select a specific translate of the striped patterns and induce algebraically decaying, dipole-type farfield deformations. Phase shifts and leading order terms are determined by effective

Roman Feller, Michael Pinsker

For a set F of finite tournaments, the F-free orientation problem is the problem of deciding if a given finite undirected graph can be oriented in such a way that the resulting oriented graph does not contain any member of F. Using the theory of smooth approximations, we give a new shorter proof of the complexity dichotomy for such problems obtained recently

Ruslan Maksimau, Catharina Stroppel

Naisse and Vaz defined an extension of KLR algebras to categorify Verma modules. We realise these algebras geometrically as convolution algebras in Borel-Moore homology. For this we introduce Grassmannian-Steinberg quiver flag varieties. They generalize Steinberg quiver flag varieties in a non-obvious way, reflecting the diagrammatics from the Naisse-Vaz con

Daniele Vignarca, Stefano Arrigoni, Edoardo Sabbioni, Federico Cheli

Intelligent Transportation Systems (ITS) are pushing an increasing interest and development when dealing with eco-driving systems. In this framework, this paper presents a method to define speed profiles specifically designed for Green Light Optimal Speed Advisory (GLOSA) systems on buses. GLOSA aims to optimize traffic flow by providing vehicles with real-t

Christian Holger Nerowski, Zongjun Li, Christian Rehtanz

The large-scale integration of distributed renewable energy sources into the electricity grid requires the investigation of new methods to ensure stability. For example, Active Distribution Networks (ADNs) can be used at (sub-) transmission levels for emergency operation, provided robust and efficient control is available. This paper investigates the use of

Ualbai Umirbaev

We show that all Chein automorphisms (or one-row transformations) of lower degree $\geq 4$ of a free metabelian Lie algebra $M_n$ of rank $n\geq 4$ over an arbitrary field $K$ of characteristic $\neq 3$ are tame. We then show that all exponential automorphisms of $M_n$ of lower degree $\geq 5$ are also tame under the same conditions. The same results hold fo

Fabio A. Faria, Mateus M. Souza, Raoni F. da S. Teixeira, Mauricio P. Segundo

The proliferation of deep learning solutions and the scarcity of large annotated datasets pose significant challenges in real-world applications. Various strategies have been explored to overcome this challenge, with data augmentation (DA) approaches emerging as prominent solutions. DA approaches involve generating additional examples by transforming existin

Daniel Weinberg, Olivia A. Hull, Jacob M. Clary, Ravishankar Sundararaman

Developing theoretical understanding of complex reactions and processes at interfaces requires using methods that go beyond semilocal density functional theory to accurately describe the interactions between solvent, reactants and substrates. Methods based on many-body perturbation theory, such as the random phase approximation (RPA), have previously been li

V. I. Murashka

L.A. Shemetkov posed a Problem 9.74 in Kourovka Notebook to find all local formations $\mathfrak{F}$ of finite groups such that every finite minimal non-$\mathfrak{F}$-group is either a Schmidt group or a group of prime order. All known solutions to this problem are obtained under the assumption that every minimal non-$\mathfrak{F}$-group is soluble. Using t

Ali Tizfahm, Saeed Fakhry, Javad T. Firouzjaee, Antonino Del Popolo

In this study, we explore gravitational lensing using more realistic dark matter halo models, moving beyond the limitations of spherical-collapse approximations. Through analytical calculations employing various mass functions, we address critical factors often neglected in the standard Press-Schechter formalism, such as ellipsoidal collapse conditions, angu

Xiaoyong Zhang, Carlos A. R. Sá de Melo

Motivated by the experimental observation of spin-charge separation in one-dimensional interacting Fermi gases, we investigate these systems in the presence of spin-orbit coupling and Rabi fields. We demonstrate that spin-charge-separated modes evolve into helical collective modes due to the special mixing of spin and charge induced by spin-orbit coupling an

Dhruv Agarwal, Farhana Shahid, Aditya Vashistha

WhatsApp groups have become a hotbed for the propagation of harmful content including misinformation, hate speech, polarizing content, and rumors, especially in Global South countries. Given the platform's end-to-end encryption, moderation responsibilities lie on group admins and members, who rarely contest such content. Another approach is fact-checking, wh

Andy Liu, Mona Diab, Daniel Fried

The task of persona-steered text generation requires large language models (LLMs) to generate text that reflects the distribution of views that an individual fitting a persona could have. People have multifaceted personas, but prior work on bias in LLM-generated opinions has only explored multiple-choice settings or one-dimensional personas. We define an inc

Yuchi Liu, Jaskirat Singh, Gaowen Liu, Ali Payani

Large language models (LLMs) have shown great progress in responding to user questions, allowing for a multitude of diverse applications. Yet, the quality of LLM outputs heavily depends on the prompt design, where a good prompt might enable the LLM to answer a very challenging question correctly. Therefore, recent works have developed many strategies for imp

Adolfo S. Carvalho, Lynne A. Hillenbrand, Jerome Seebeck, Kevin Covey

We present a comprehensive analysis of the post-outburst evolution of the FU Ori object HBC 722 in optical/near-infrared (NIR) photometry and spectroscopy. Using a modified viscous accretion disk model, we fit the outburst epoch SED to determine the physical parameters of the disk, including $\dot{M}_\mathrm{acc} = 10^{-4.0} \ M_\odot$ yr$^{-1}$, $R_\mathrm{

Deven Sethi, David Šiška, Yufei Zhang

Entropy regularization has been widely used in policy optimization algorithms to enhance exploration and the robustness of the optimal control; however it also introduces an additional regularization bias. This work quantifies the impact of entropy regularization on the convergence of policy gradient methods for stochastic exit time control problems. We anal

A. Alonso-Izquierdo, N. S. Manton, J. Mateos Guilarte, A. Wereszczynski

Models are developed for the motion of charge-2 Abelian Higgs vortices through the 2-vortex moduli space $M$, with the vortices excited by their shape mode oscillations. The models simplify to the well-known geodesic flow on $M$, modified by a potential, when the mode oscillations are fast relative to the moduli space motion and their amplitudes are small. W