The use of machine learning (ML) in health care raises numerous ethicalconcerns, especially as models can amplify existing health inequities. Here, weoutline ethical considerations for equitable... | Continue reading
Online social networks provide a platform for sharing information and freeexpression. However, these networks are also used for malicious purposes, suchas distributing misinformation and hate... | Continue reading
Index structures are a building block of query processing and computerscience in general. Since the dawn of computer technology there have been indexstructures. And since then, a myriad of index... | Continue reading
In this work we target the problem of provably computing the equivalencebetween two programs represented as dataflow graphs. To this end, we formalizethe problem of equivalence between two... | Continue reading
We consider the paradigm of a black box AI system that makes life-criticaldecisions. We propose an "arguing machines" framework that pairs the primary AIsystem with a secondary one that is... | Continue reading
Recent work has explored the possibility of pruning neural networks atinitialization. We assess proposals for doing so: SNIP (Lee et al., 2019),GraSP (Wang et al., 2020), SynFlow (Tanaka et al.,... | Continue reading
Mitigating the risk arising from extreme events is a fundamental goal withmany applications, such as the modelling of natural disasters, financialcrashes, epidemics, and many others. To manage... | Continue reading
New machine learning accelerators are being announced and released each monthfor a variety of applications from speech recognition, video object detection,assisted driving, and many data center... | Continue reading
When scaled to hundreds of billions of parameters, pretrained language modelssuch as GPT-3 (Brown et al., 2020) achieve remarkable few-shot performance onchallenging natural language... | Continue reading
Recent progress in Game AI has demonstrated that given enough data from humangameplay, or experience gained via simulations, machines can rival or surpassthe most skilled human players in... | Continue reading
Quantum mechanics is derived as an application of the method of maximumentropy. No appeal is made to any underlying classical action principle whetherdeterministic or stochastic. Instead, the... | Continue reading
Automated game design is the problem of automatically producing games throughcomputational processes. Traditionally these methods have relied on theauthoring of search spaces by a designer,... | Continue reading
Normalizing flows, autoregressive models, variational autoencoders (VAEs),and deep energy-based models are among competing likelihood-based frameworksfor deep generative learning. Among them,... | Continue reading
Given a machine $U$, a $c$-short program for $x$ is a string $p$ such that$U(p)=x$ and the length of $p$ is bounded by $c$ + (the length of a shortestprogram for $x$). We show that for any... | Continue reading
Recent work in deep reinforcement learning (RL) has produced algorithmscapable of mastering challenging games such as Go, chess, or shogi. In theseworks the RL agent directly observes the... | Continue reading
There has been a recent media blitz on a cohort of mathematicians valiantlyworking to fix America's democratic system by combatting gerrymandering withgeometry. While statistics commonly... | Continue reading
Brain decoding, understood as the process of mapping brain activities to thestimuli that generated them, has been an active research area in the lastyears. In the case of language stimuli,... | Continue reading
Hardware, systems and algorithms research communities have historically haddifferent incentive structures and fluctuating motivation to engage with eachother explicitly. This historical... | Continue reading
Spectre and Meltdown attacks and their variants exploit performanceoptimization features to cause security breaches. Secret information isaccessed and leaked through micro-architectural covert... | Continue reading
The putative detection of phosphine in the atmosphere of Venus at anabundance of $\sim 20$ ppb suggests that this gas is being generated by eitherindeterminate abiotic pathways or biological... | Continue reading
We derive the backpropagation algorithm for spiking neural networks composedof leaky integrate-and-fire neurons operating in continuous time. Thisalgorithm, EventProp, computes the exact... | Continue reading
Transformer model architectures have garnered immense interest lately due totheir effectiveness across a range of domains like language, vision andreinforcement learning. In the field of natural... | Continue reading
Hardware, systems and algorithms research communities have historically haddifferent incentive structures and fluctuating motivation to engage with eachother explicitly. This historical... | Continue reading
We show that an equal-mass, temporary binary companion to the Sun in thesolar birth cluster at a separation of $\sim 10^3 \; \mathrm{\; AU}$ would haveincreased the likelihood of forming the... | Continue reading
Virtual memory has been a standard hardware feature for more than threedecades. At the price of increased hardware complexity, it has simplifiedsoftware and promised strong isolation among... | Continue reading
Forecasting elections -- a challenging, high-stakes problem -- is the subjectof much uncertainty, subjectivity, and media scrutiny. To shed light on thisprocess, we develop a method for... | Continue reading
Quantum algorithms could be much faster than classical ones in solving thefactoring problem. Adiabatic quantum computation for this is an alternativeapproach other than Shor's algorithm. Here we... | Continue reading
We consider how to forecast progress in the domain of quantum computing. Forthis purpose we collect a dataset of quantum computer systems to date, scoredon their physical qubits and gate error... | Continue reading
The COVID-19 pandemic has shaken the world to its core and has provoked anovernight exodus of developers that normally worked in an office setting toworking from home. The magnitude of this... | Continue reading
A critical issue in software development projects in IT service companies isfinding the right people at the right time. By enabling assignments of tasks topeople to be more fluid, the use of... | Continue reading
Interprocess communication, IPC, is one of the most fundamental functions ofa modern operating system, playing an essential role in the fabric ofcontemporary applications. This report conducts... | Continue reading
When absorbing boundary conditions are used to evaporate a black hole inAdS/CFT, we show that there is a phase transition in the location of thequantum Ryu-Takayanagi surface, at precisely the... | Continue reading
We compare the impact of hardware advancement and algorithm advancement forSAT solving over the last two decades. In particular, we compare 20-year-oldSAT-solvers on new computer hardware with... | Continue reading
These are notes for a graduate course in quantum gravity planned atIFT-UAM/CSIC for the spring term of 2020, but delayed for an indefinite periodof time (although a shortened version was given... | Continue reading
Causality visualization can help people understand temporal chains of events,such as messages sent in a distributed system, cause and effect in a historicalconflict, or the interplay between... | Continue reading
Motivated by the fast adoption of WebAssembly, we propose the firstfunctional pipeline to support the superoptimization of WebAssembly bytecode.Our pipeline works over LLVM and Souper. We... | Continue reading
We argue, in the context of Ads/CFT correspondence, that the degree ofentanglement on the CFTs side determines the orientation of space and time onthe dual global spacetime. That is, the global... | Continue reading
(abridged) Both fundamental constants that describe the laws of physics andcosmological parameters that determine the cosmic properties must fall within arange of values in order for the... | Continue reading
This paper introduces WaveGrad, a conditional model for waveform generationthrough estimating gradients of the data density. This model is built on theprior work on score matching and diffusion... | Continue reading
It is non-trivial to design engaging and balanced sets of game rules. Modernchess has evolved over centuries, but without a similar recourse to history,the consequences of rule changes to game... | Continue reading
Challenging problems such as open-domain question answering, fact checking,slot filling and entity linking require access to large, external knowledgesources. While some models do well on... | Continue reading
The COVID-19 pandemic has had a wide-ranging impact on information workerssuch as higher stress levels, increased workloads, new workstreams, and morecaregiving responsibilities during lockdown.... | Continue reading
We propose a new test to measure a text model's multitask accuracy. The testcovers 57 tasks including elementary mathematics, US history, computer science,law, and more. To attain high accuracy... | Continue reading
We encode the sequence of prime numbers into simple superpositions ofidentical waves, mimicking the archetypal prime number sieve of Eratosthenes.The primes are identified as zeros accompanied... | Continue reading
We explore the application of transformer-based language models to automatedtheorem proving. This work is motivated by the possibility that a majorlimitation of automated theorem provers... | Continue reading
Animated transitions help viewers follow changes between relatedvisualizations. Specifying effective animations demands significant effort:authors must select the elements and properties to... | Continue reading
Most current self-sovereign identity systems may be categorized as strictlyobjective, consisting of cryptographically signed statements issued by trustedthird party attestors. This failure to... | Continue reading