One of the major optimizations employed in deep learning frameworks is graphrewriting. Production frameworks rely on heuristics to decide if rewrite rulesshould be applied and in which order.... | Continue reading
A large-scale, up-to-date analysis of Computer Science literature (11.8Mpapers through 2019) reveals that, if trends from the last 50 years continue,parity between the number of male and female... | Continue reading
This book is an introduction to program analysis that is meant to beconsiderably more elementary than our advanced book Principles of ProgramAnalysis (Springer, 2005). Rather than using flow... | Continue reading
The SIMT execution model is commonly used for general GPU development. CUDAand OpenCL developers write scalar code that is implicitly parallelized bycompiler and hardware. On Intel GPUs,... | Continue reading
Cloud gaming is a new class of services that promises to revolutionize thevideogame market. It allows the user to play a videogame with basic equipmentwhile using a remote server for the actual... | Continue reading
It has been observed that large-scale language models capture undesirablesocietal biases, e.g. relating to race and gender; yet religious bias has beenrelatively unexplored. We demonstrate that... | Continue reading
The last few years have seen gigantic leaps in algorithms and systems tosupport efficient deep learning inference. Pruning and quantization algorithmscan now consistently compress neural... | Continue reading
Members of high performing software teams collaborate, exchange informationand coordinate their work on a frequent, regular basis. Most teams have thedaily stand-up meeting as a central venue... | Continue reading
Counters are an important abstraction in distributed computing, and play acentral role in large scale geo-replicated systems, counting events such as webpage impressions or social network... | Continue reading
This paper presents Riposte, a new system for anonymous broadcast messaging.Riposte is the first such system, to our knowledge, that simultaneouslyprotects against traffic-analysis attacks,... | Continue reading
This paper provides a self-contained first introduction to description logics(DLs). The main concepts and features are explained with examples before syntaxand semantics of the DL SROIQ are... | Continue reading
In recent years, Ethereum gained tremendously in popularity, growing from adaily transaction average of 10K in January 2016 to an average of 500K inJanuary 2020. Similarly, smart contracts began... | Continue reading
Despite the increasingly important role played by image memes, we do not yethave a solid understanding of the elements that might make a meme go viral onsocial media. In this paper, we... | Continue reading
Program tracing, or mentally simulating a program on concrete inputs, is animportant part of general program comprehension. Programs involve many kinds ofvirtual state that must be held in... | Continue reading
Attributing a piece of malware to its creator typically requires threatintelligence. Binary attribution increases the level of difficulty as it mostlyrelies upon the ability to disassemble... | Continue reading
Flat rotation curves in disk galaxies represent the main evidence for largeamounts of surrounding dark matter. Despite of the difficulty in identifyingthe dark matter contribution to the total... | Continue reading
We propose a method for meta-learning reinforcement learning algorithms bysearching over the space of computational graphs which compute the lossfunction for a value-based model-free RL agent to... | Continue reading
Credit allows a lender to loan out surplus capital to a borrower. In thetraditional economy, credit bears the risk that the borrower may default on itsdebt, the lender hence requires an upfront... | Continue reading
We use a contrastive self-supervised learning framework to estimate distancesto galaxies from their photometric images. We incorporate data augmentationsfrom computer vision as well as an... | Continue reading
Background: Test-driven development (TDD) is a technique that repeats shortcoding cycles interleaved with testing. The developer first writes a unit testfor the desired functionality, followed... | Continue reading
Recent works in Grammatical Error Correction (GEC) have leveraged theprogress in Neural Machine Translation (NMT), to learn rewrites from parallelcorpora of grammatically incorrect and corrected... | Continue reading
Recent work on Grammatical Error Correction (GEC) has highlighted theimportance of language modeling in that it is certainly possible to achievegood performance by comparing the probabilities of... | Continue reading
Parler is as an alternative social network promoting itself as a service thatallows its users to "Speak freely and express yourself openly, without fear ofbeing deplatformed for your views."... | Continue reading
Nearly twenty years after the launch of AWS, it remains difficult for mostdevelopers to harness the enormous potential of the cloud. In this paper we layout an agenda for a new generation of... | Continue reading
In deep learning, models typically reuse the same parameters for all inputs.Mixture of Experts (MoE) defies this and instead selects different parametersfor each incoming example. The result is... | Continue reading
Political trolls initiate online discord not only for the lulz (laughs) butalso for ideological reasons, such as promoting their desired politicalcandidates. Political troll groups recently... | Continue reading
Network science is the field dedicated to the investigation and analysis ofcomplex systems via their representations as networks. We normally model suchnetworks as graphs: sets of nodes... | Continue reading
This paper describes PALMED, a tool that automatically builds a resourcemapping, a performance model for pipelined, super-scalar, out-of-order CPUarchitectures. Resource mappings describe the... | Continue reading
We re-examine the constraints imposed by causality and unitarity on thelow-energy effective field theory expansion of four-particle scatteringamplitudes, exposing a hidden "totally positive"... | Continue reading
John Horton Conway's Cosmological Theorem, about Audioactive sequences, forwhich no extant proof existed, is given a computer-generated proof, hopefullyfor good. | Continue reading
We apply methods of topological data analysis to loss functions to gaininsights on learning of deep neural networks and their generalizationproperties. We study global properties of the loss... | Continue reading
We propose CHARM, a method for training a single neural network acrossinconsistent input channels. Our work is motivated by Electroencephalography(EEG), where data collection protocols from... | Continue reading
Algebraic quantum field theory and prefactorization algebra are twomathematical approaches to quantum field theory. In this monograph, using a newcoend definition of the Boardman-Vogt... | Continue reading
We tackle the problem of machine translation of manga, Japanese comics. Mangatranslation involves two important problems in machine translation:context-aware and multimodal translation. Since... | Continue reading
Voice assistants are now ubiquitous and listen in on our everyday lives. Eversince they became commercially available, privacy advocates worried that thedata they collect can be abused: might... | Continue reading
This paper presents {\scshape PaSh}, a system for parallelizing POSIX shellscripts. Given a script, {\scshape PaSh} converts it to a dataflow graph,performs a series of semantics-preserving... | Continue reading
A key limitation in using various modern methods of machine learning indeveloping feedback control policies is the lack of appropriate methodologiesto analyze their long-term dynamics, in terms... | Continue reading
In the field of Astrobiology, the precise location, prevalence and age ofpotential extraterrestrial intelligence (ETI) have not been explicitlyexplored. Here, we address these inquiries using an... | Continue reading
Future spacecraft will require a paradigm shift in the way the information istransmitted due to the continuous increase in the amount of data requiringspace links. Current radiofrequency-based... | Continue reading
This paper proposes a research direction to advance AI which drawsinspiration from cognitive theories of human decision making. The premise isthat if we gain insights about the causes of some... | Continue reading
Deep learning is a group of exciting new technologies for neural networks.Through a combination of advanced training techniques and neural networkarchitectural components, it is now possible to... | Continue reading
The ongoing commercial deployment of millimeter wave systems brings intofocus a number of practical issues in form factor user equipment (UE) design.With wavelengths becoming smaller, antenna... | Continue reading
We introduce a new approach to deep learning on 3D surfaces such as meshes orpoint clouds. Our key insight is that a simple learned diffusion layer canspatially share data in a principled... | Continue reading
Session types statically describe communication protocols between concurrentmessage-passing processes. Unfortunately, parametric polymorphism even in itsrestricted prenex form is not fully... | Continue reading
The Lightning Network promises to alleviate Bitcoin's known scalabilityproblems. The operation of such second layer approaches relies on the abilityof participants to turn to the blockchain to... | Continue reading
Graded type theories are an emerging paradigm for augmenting the reasoningpower of types with parameterizable, fine-grained analyses of programproperties. There have been many such theories in... | Continue reading
Consistent hashing (CH) is a central building block in many networkingapplications, from datacenter load-balancing to distributed storage.Unfortunately, state-of-the-art CH solutions cannot... | Continue reading
Face detection has received intensive attention in recent years. Many workspresent lots of special methods for face detection from different perspectiveslike model architecture, data... | Continue reading