Modern cloud services are designed to encourage and support collaboration. Tohelp users share links to online documents, maps, etc., several services,including cloud storage providers such as... | Continue reading
The conventional framework for defining and understanding phases of matterrequires thermodynamic equilibrium. Extensions to non-equilibrium systems haveled to surprising insights into the nature... | Continue reading
A key element of AutoML systems is setting the types of models that will beused for each type of task. For classification and regression problems withtabular data, the use of tree ensemble... | Continue reading
Neural radiance fields (NeRF) methods have demonstrated impressive novel viewsynthesis performance. The core approach is to render individual rays byquerying a neural network at points sampled... | Continue reading
Implicit representations of geometry, such as occupancy fields or signeddistance fields (SDF), have recently re-gained popularity in encoding 3D solidshape in a functional form. In this work, we... | Continue reading
Three factors drive the advance of AI: algorithmic innovation, data, and theamount of compute available for training. Algorithmic progress hastraditionally been more difficult to quantify than... | Continue reading
We describe the public release of the Cluster Monte Carlo Code (CMC) aparallel, star-by-star $N$-body code for modeling dense star clusters. CMCtreats collisional stellar dynamics using... | Continue reading
We introduce a new timing side-channel attack on Intel CPU processors. OurFrontal attack exploits timing differences that arise from how the CPU frontendfetches and processes instructions while... | Continue reading
In this paper, two novel automatic and real-time systems for the detectionand classification of two genera of scorpions found in La Plata city(Argentina) were developed using computer vision and... | Continue reading
The DNS filtering apparatus of China's Great Firewall (GFW) has evolvedconsiderably over the past two decades. However, most prior studies of China'sDNS filtering were performed over short time... | Continue reading
In the last three years, the largest dense deep learning models have grownover 1000x to reach hundreds of billions of parameters, while the GPU memoryhas only grown by 5x (16 GB to 80 GB).... | Continue reading
We develop a fast and accurate algorithm for evaluating $\sqrt{a^2+b^2}$ fortwo floating point numbers $a$ and $b$. Library functions that perform thiscomputation are generally named {\tt... | Continue reading
We present a source-to-source transformation framework, GOCC, that consumeslock-based pessimistic concurrency programs in the Go language and transformsthem into optimistic concurrency programs... | Continue reading
Programs written in C/C++ can suffer from serious memory fragmentation,leading to low utilization of memory, degraded performance, and applicationfailure due to memory exhaustion. This paper... | Continue reading
Mixture-of-Experts (MoE) models can achieve promising results with outrageouslarge amount of parameters but constant computation cost, and thus it hasbecome a trend in model scaling. Still it is... | Continue reading
Intersecting neuroscience and deep learning has brought benefits anddevelopments to both fields for several decades, which help to both understandhow learning works in the brain, and to achieve... | Continue reading
We explore the Collatz conjecture and its variants through the lens oftermination of string rewriting. We construct a rewriting system that simulatesthe iterated application of the Collatz... | Continue reading
Neuromorphic computing takes inspiration from the brain to create energyefficient hardware for information processing, capable of highly sophisticatedtasks. In this article, we make the case... | Continue reading
In this article for the first time, comprehensive studies of mosquitoneutralization using machine vision and a 1 W power laser are considered.Developed laser installation with Raspberry Pi that... | Continue reading
Earth-like planets are expected to provide the greatest opportunity for thedetection of life beyond the Solar System. However our planet cannot beconsidered a fair sample, especially if... | Continue reading
Debugging is an essential process with a large share of the developmenteffort, being a relentless quest for offensive code through tracing, inspectionand iterative running sessions. Probably... | Continue reading
Personalized recommendations on the Netflix Homepage are based on a user'sviewing habits and the behavior of similar users. These recommendations,organized for efficient browsing, enable users... | Continue reading
To evaluate the role of Total Solar Irradiance (TSI) on Northern Hemisphere(NH) surface air temperature trends it is important to have reliable estimatesof both quantities. 16 different TSI... | Continue reading
Inductive logic programming (ILP) is a form of machine learning. The goal ofILP is to induce a hypothesis (a set of logical rules) that generalises giventraining examples. In contrast to most... | Continue reading
Face recognition has achieved significant progress in deep-learning era dueto the ultra-large-scale and well-labeled datasets. However, training onultra-large-scale datasets is time-consuming... | Continue reading
Recent work has demonstrated that increased training dataset diversityimproves general cross-domain knowledge and downstream generalizationcapability for large-scale language models. With this... | Continue reading
Text-to-Image generation in the general domain has long been an open problem,which requires both generative model and cross-modal understanding. We proposeCogView, a 4-billion-parameter... | Continue reading
The distribution of entangled states across the nodes of a future quantuminternet will unlock fundamentally new technologies. Here we report on theexperimental realization of a three-node... | Continue reading
With the increased availability of rich tactile sensors, there is an equallyproportional need for open-source and integrated software capable ofefficiently and effectively processing raw touch... | Continue reading
For all the ways convolutional neural nets have revolutionized computervision in recent years, one important aspect has received surprisingly littleattention: the effect of image size on the... | Continue reading
Threats to passwords are still very relevant due to attacks like phishing orcredential stuffing. One way to solve this problem is to remove passwordscompletely. User studies on passwordless... | Continue reading
We present the first proof-of-concept simulations of detectors usingbiomaterials to detect particle interactions. The essential idea behind a "DNAdetector" involves the attachment of a forest of... | Continue reading
We propose a design for philanthropic or publicly-funded seeding to allow(near) optimal provision of a decentralized, self-organizing ecosystem ofpublic goods. The concept extends ideas from... | Continue reading
Existing image-to-image translation (I2IT) methods are either constrained tolow-resolution images or long inference time due to their heavy computationalburden on the convolution of... | Continue reading
People ``understand'' the world via vision, hearing, tactile, and also thepast experience. Human experience can be learned through normal learning (wecall it explicit knowledge), or... | Continue reading
Decomposing knowledge into interchangeable pieces promises a generalizationadvantage when there are changes in distribution. A learning agent interactingwith its environment is likely to be... | Continue reading
Let $A$ be drawn uniformly at random from the set of all $n\times n$symmetric matrices with entries in $\{-1,1\}$. We show that \[ \mathbb{P}(\det(A) = 0 ) \leq e^{-cn},\] where $c>0$ is an... | Continue reading
Technologies play an important role in the hiring process for softwareprofessionals. Within this process, several studies revealed misconceptions andbad practices which lead to suboptimal... | Continue reading
While programming is one of the most broadly applicable skills in modernsociety, modern machine learning models still cannot code solutions to basicproblems. It can be difficult to accurately... | Continue reading
This paper presents Tree Notation, a new simple, universal syntax. Languagedesigners can invent new programming languages, called Tree Languages, on topof Tree Notation. Tree Languages have a... | Continue reading
The advent of Generative Adversarial Network (GAN) architectures has givenanyone the ability of generating incredibly realistic synthetic imagery. Themalicious diffusion of GAN-generated images... | Continue reading
RISC-V is a promising open-source architecture primarily targeted forembedded systems. Programs compiled using the RISC-V toolchain can runbare-metal on the system, and, as such, can be... | Continue reading
This work attempts to provide a plausible theoretical framework that aims tointerpret modern deep (convolutional) networks from the principles of datacompression and discriminative... | Continue reading
Vision transformers (ViT) have demonstrated impressive performance acrossvarious machine vision problems. These models are based on multi-headself-attention mechanisms that can flexibly attend... | Continue reading
Transformers have become one of the most important architectural innovationsin deep learning and have enabled many breakthroughs over the past few years.Here we propose a simple attention-free... | Continue reading
Nowadays, in operating systems, numerous protection mechanisms prevent orlimit the user-mode applicationsto access the kernels internal information.This is regularlycarried out by software-based... | Continue reading
This paper introduces a generative model equivariant to Euclidean symmetries:E(n) Equivariant Normalizing Flows (E-NFs). To construct E-NFs, we take thediscriminative E(n) graph neural networks... | Continue reading
The second quantum technological revolution started around 1980 with thecontrol of single quantum particles and their interaction on an individualbasis. These experimental achievements enabled... | Continue reading