24 papers · ranked by Valyu relevance
Ye Lin, Mingxuan Wang, Zhexi Zhang, Xiaohui Wang + 2 more
'Jingbo Zhu'] Parameter sharing has proven to be a parameterefficient approach. Previous work on Transformers has focused on sharing parameters in different layers, which can improve the performance of models with limited parameters by increasing model depth. In this paper, we study why this approach works from two…
Biao Zhao, Weiqiang Jin, Zhang Chen, Yucheng Guo
Humans do not learn everything from the scratch but can connect and associate the upcoming information with the exchanged experience and known knowledge. Such an idea can be extended to cooperated multi-reinforcement learning and has achieved its success on homogeneous agents by means of parameter sharing. However, it…
Filippos Christianos, Georgios Papoudakis, Muhammad Arrasy Rahman, Stefano V. Albrecht
'Stefano V. Albrecht'] Sharing parameters in multi-agent deep reinforcement learning has played an essential role in allowing algorithms to scale to a large number of agents. Parameter sharing between agents significantly decreases the number of trainable parameters, shortening training times to tractable levels, and…
Avanti Shrikumar, Peyton Greenside, Anshul Kundaje
Deep learning approaches that have produced breakthrough predictive models in computer vision, speech recognition and machine translation are now being successfully applied to problems in regulatory genomics. However, deep learning architectures used thus far in genomics are often directly ported from computer vision…
Viktor Martinek, Roland W. Herzog
Symbolic regression aims to find symbolic expressions that describe datasets. Due to better interpretability, it is a machine learning paradigm particularly powerful for scientific discovery. In recent years, several works have expanded the concept to allow the description of similar phenomena using a single expression…
Ying Chen, Jiong Yu, Yutong Zhao, Jiaying Chen + 4 more
'Andrea Prati' 'Luis Javier García Villalba' 'Vincent A. Cicirello'] In most of the existing multi-task learning (MTL) models, multiple tasks’ public information is learned by sharing parameters across hidden layers, such as hard sharing, soft sharing, and hierarchical sharing. One promising approach is to introduce…
Imen Bouhlel, Charley M. Wu, Nobuyuki Hanaki, Robert L. Goldstone
Information sharing in competitive environments may seem counterintuitive, yet it is widely observed in humans and other animals. For instance, the open-source software movement has led to new and valuable technologies being released publicly to facilitate broader collaboration and further innovation. What drives this…
Chanseok Park, Min Wang, Refah Mohammed Alotaibi, Hoda Rezk
A load-sharing system is defined as a parallel system whose load will be redistributed to its surviving components as each of the components fails in the system. Our focus is on making statistical inference of the parameters associated with the lifetime distribution of each component in the system. In this paper, we…
Hervé Paulino, Eduardo R. B. Marques
Heterogeneity is omnipresent in today's commodity computational systems, which comprise at least one multi-core Central Processing Unit (CPU) and one Graphics Processing Unit (GPU). Nonetheless, all this computing power is not being harnessed in mainstream computing, as the programming of these systems entails many…
Nicholas Menghi, Kemal Kacar, Will Penny
This paper uses constructs from the field of multitask machine learning to define pairs of learning tasks that either shared or did not share a common subspace. Human subjects then learnt these tasks using a feedback-based approach. We found, as hypothesised, that subject performance was significantly higher on the…
Arthur S. Powanwe, Andre Longtin, Cinzia Costa
Simple Summary To properly interact with our environment, the brain must be able to identify external stimuli, process them, and make the right decisions all in a short time. This may involve several brain regions interacting together by sharing information birectionally via rhythmic activity. Such flexibility requires…
Authors not listed
The discoverability and reusability of data is critical for machine learning to drive new discovery in the chemical sciences, and the ‘FAIR Guiding Principles for scientific data management and stewardship’ provide a measurable set of guidelines that can be used to ensure the accessibility of reusable data. We…
Gheorghe-Teodor Bercea, Carlo Bertolli, Arpith C. Jacob, Alexandre E. Eichenberger + 5 more
'Alexandre E. Eichenberger' 'Alexey Bataev' 'Georgios Rokos' 'Hyojin Sung' 'Tong Chen' 'Kevin O’Brien'] OpenMP is a shared memory programming model which supports the offloading of target regions to accelerators such as NVIDIA GPUs. e implementation in Clang/LLVM aims to deliver a generic GPU compilation toolchain that…
Leif Jacobson, James Stevenson, Farhad Ramezanghorbani, Steven Dajnowicz + 1 more
Transferable neural network potentials have shown great promise as an avenue to increase the accuracy and applicability of existing atomistic force fields for organic molecules and inorganic materials. Training sets used to develop transferable potentials are very large, typically millions of examples, and as such, are…
Pieter Verbeke, Tom Verguts
Human adaptive behavior requires continually learning and performing a wide variety of tasks, often with very little practice. To accomplish this, it is crucial to separate neural representations of different tasks in order to avoid interference. At the same time, sharing neural representations supports generalization…
David Algis, Bérenger Bramas, Emmanuelle Darles, Lilian Aveneau
and Few Particles per Cell Authors: ['David Algis' 'Bérenger Bramas' 'Emmanuelle Darles' 'Lilian Aveneau'] This paper presents novel approaches to parallelizing particle interactions on a GPU when there are few particles per cell and the interactions are limited by a cutoff distance. The paper surveys classical…
Nora G. Peterson, Benjamin M. Stormo, Kevin P. Schoenfelder, Juliet S. King + 2 more
Multiple nuclei sharing a common cytoplasm are found in diverse tissues, organisms, and diseases. Yet, multinucleation remains a poorly understood biological property. Cytoplasm sharing invariably involves plasma membrane breaches. In contrast, we discovered cytoplasm sharing without membrane breaching in highly…
Authors not listed
The first step in discovering novel medicines is identifying an initial hit against a biological target or organism, followed by extensive iterative processes to explore the structure-activity relationship and establish a lead compound. As a result, drug discovery laboratories tend to accumulate extensive collections…
Inés P. Mariño, Ekkehard Ullner, Alexey Zaikin, Jürgen Kurths
We have investigated simulation-based techniques for parameter estimation in chaotic intercellular networks. The proposed methodology combines a synchronization-based framework for parameter estimation in coupled chaotic systems with some state-of-the-art computational inference methods borrowed from the field of…
Joshua Horton, Simon Boothroyd, Jeffrey Wagner, Joshua Mitchell + 9 more
The development of accurate transferable force fields is key to realizing the full potential of atomistic modelling in the study of biological processes such as protein--ligand binding for drug discovery. State-of-the-art transferable force fields, such as those produced by the Open Force Field Initiative, use modern…
Farhan Musanna, Sanjeev Kumar
In this work, we present a quantum secret sharing scheme based on Bell state entanglement and sequential projection measurements. The protocol verifies the n out of n scheme and supports the aborting of the protocol in case all the parties do not divulge in their valid measurement outcomes. The operator-qubit pair…
Andrew Teale, Trygve Helgaker, Andreas Savin, Carlo Adamo + 66 more
In this paper, the history, present status, and future of density-functional theory (DFT) is informally reviewed and discussed by 70 workers in the field, including molecular scientists, materials scientists, method developers and practitioners. The format of the paper is that of a roundtable discussion, in which the…
Riley Hickman, Priyansh Parakh, Austin Cheng, Qianxiang Ai + 3 more
Experiment planning algorithms are a required component of autonomous platforms for scientific discovery. Selecting a suitable optimization algorithm for a novel application is an important yet difficult choice a researcher has to make based on past empirical performance on similar tasks. To facilitate the evaluation…
Authors not listed
Machine olfaction—the artificial replication of the sense of smell—faces significant challenges due to the absence of large, standardized training datasets. Unlike vision, language, and audio models, which benefit from extensive corpora such as ImageNet, GLUE, and AudioSet, olfaction lacks scaled equivalents and…