Search Results - "Shaham, Uri"

Understanding adversarial training: Increasing local stability of supervised models through robust optimization by Shaham, Uri, Yamada, Yutaro, Negahban, Sahand

“…We show that adversarial training of supervised learning models is in fact a robust optimization procedure. To do this, we establish a general framework for…”
Get full text

DeepSurv: personalized treatment recommender system using a Cox proportional hazards deep neural network by Katzman, Jared L, Shaham, Uri, Cloninger, Alexander, Bates, Jonathan, Jiang, Tingting, Kluger, Yuval

“…Medical practitioners use survival models to explore and understand the relationships between patients' covariates (e.g. clinical and genetic features) and the…”
Get full text

Learning by coincidence: Siamese networks and common variable learning by Shaham, Uri, Lederman, Roy R.

“…•Siamese networks are proposed for unsupervised learning of a common source of variability in multimodal data.•The common source of variability is identified…”
Get full text

Efficient Long-Text Understanding with Short-Text Models by Ivgi, Maor, Shaham, Uri, Berant, Jonathan

“…Transformer-based pretrained language models (LMs) are ubiquitous across natural language understanding, but cannot be applied to long sequences such as…”
Get full text

Removal of batch effects using distribution-matching residual networks by Shaham, Uri, Stanton, Kelly P, Zhao, Jun, Li, Huamin, Raddassi, Khadir, Montgomery, Ruth, Kluger, Yuval

“…Sources of variability in experimentally derived data include measurement error in addition to the physical phenomena of interest. This measurement error is a…”
Get full text

Gating mass cytometry data by deep learning by Li, Huamin, Shaham, Uri, Stanton, Kelly P, Yao, Yi, Montgomery, Ruth R, Kluger, Yuval

“…Mass cytometry or CyTOF is an emerging technology for high-dimensional multiparameter single cell analysis that overcomes many limitations of…”
Get full text

Deep unsupervised feature selection by discarding nuisance and correlated features by Shaham, Uri, Lindenbaum, Ofir, Svirsky, Jonathan, Kluger, Yuval

“…Modern datasets often contain large subsets of correlated features and nuisance features, which are not or loosely related to the main underlying structures of…”
Get full text

Algorithms, Applications and Theoretical Properties of Deep Neural Networks by Shaham, Uri

“…Various theoretical and practical aspects of deep learning are discussed. We present approximation bounds for deep neural networks and their dependence on the…”
Get full text

SpecRaGE: Robust and Generalizable Multi-view Spectral Representation Learning by Yacobi, Amitai, Lindenbaum, Ofir, Shaham, Uri

“…Multi-view representation learning (MvRL) has garnered substantial attention in recent years, driven by the increasing demand for applications that can…”
Get full text

G-SPARC: SPectral ARchitectures tackling the Cold-start problem in Graph learning by Jacobs, Yahel, Dayan, Reut, Shaham, Uri

“…Graphs play a central role in modeling complex relationships across various domains. Most graph learning methods rely heavily on neighborhood information,…”
Get full text

What Do You Get When You Cross Beam Search with Nucleus Sampling? by Shaham, Uri, Levy, Omer

“…We combine beam search with the probabilistic pruning technique of nucleus sampling to create two deterministic nucleus search algorithms for natural language…”
Get full text

Sparse Binarization for Fast Keyword Spotting by Svirsky, Jonathan, Shaham, Uri, Lindenbaum, Ofir

“…With the increasing prevalence of voice-activated devices and applications, keyword spotting (KWS) models enable users to interact with technology hands-free,…”
Get full text

Neural Machine Translation without Embeddings by Shaham, Uri, Levy, Omer

“…Many NLP models operate over sequences of subword tokens produced by hand-crafted tokenization rules and heuristic subword induction algorithms. A simple…”
Get full text

Efficient Long-Text Understanding with Short-Text Models by Ivgi, Maor, Shaham, Uri, Berant, Jonathan

“…Transformer-based pretrained language models (LMs) are ubiquitous across natural language understanding, but cannot be applied to long sequences such as…”
Get full text

ZeroSCROLLS: A Zero-Shot Benchmark for Long Text Understanding by Shaham, Uri, Ivgi, Maor, Efrat, Avia, Berant, Jonathan, Levy, Omer

“…We introduce ZeroSCROLLS, a zero-shot benchmark for natural language understanding over long texts, which contains only test and small validation sets, without…”
Get full text

Causes and Cures for Interference in Multilingual Translation by Shaham, Uri, Elbayad, Maha, Goswami, Vedanuj, Levy, Omer, Bhosale, Shruti

“…Multilingual machine translation models can benefit from synergy between different language pairs, but also suffer from interference. While there is a growing…”
Get full text

Deep Ordinal Regression using Optimal Transport Loss and Unimodal Output Probabilities by Shaham, Uri, Zaidman, Igal, Svirsky, Jonathan

“…It is often desired that ordinal regression models yield unimodal predictions. However, in many recent works this characteristic is either absent, or…”
Get full text

Instruction Induction: From Few Examples to Natural Language Task Descriptions by Honovich, Or, Shaham, Uri, Bowman, Samuel R, Levy, Omer

“…Large language models are able to perform a task by conditioning on a few input-output demonstrations - a paradigm known as in-context learning. We show that…”
Get full text

Multilingual Instruction Tuning With Just a Pinch of Multilinguality by Shaham, Uri, Herzig, Jonathan, Aharoni, Roee, Szpektor, Idan, Tsarfaty, Reut, Eyal, Matan

“…As instruction-tuned large language models (LLMs) gain global adoption, their ability to follow instructions in multiple languages becomes increasingly…”
Get full text

CoverBench: A Challenging Benchmark for Complex Claim Verification by Jacovi, Alon, Ambar, Moran, Ben-David, Eyal, Shaham, Uri, Feder, Amir, Geva, Mor, Marcus, Dror, Caciularu, Avi

“…There is a growing line of research on verifying the correctness of language models' outputs. At the same time, LMs are being used to tackle complex queries…”
Get full text