Oren Pereg
YOU?
Author Swipe
View article: Out-of-Vocabulary Sampling Boosts Speculative Decoding
Out-of-Vocabulary Sampling Boosts Speculative Decoding Open
Speculative decoding relies on fast and accurate drafters. Recent state-of-the-art language models employ larger and larger vocabularies, which significantly slows down drafters. One promising approach to boost the efficiency of speculativ…
View article: Accelerating LLM Inference with Lossless Speculative Decoding Algorithms for Heterogeneous Vocabularies
Accelerating LLM Inference with Lossless Speculative Decoding Algorithms for Heterogeneous Vocabularies Open
Accelerating the inference of large language models (LLMs) is a critical challenge in generative AI. Speculative decoding (SD) methods offer substantial efficiency gains by generating multiple tokens using a single target forward pass. How…
View article: Distributed Speculative Inference (DSI): Speculation Parallelism for Provably Faster Lossless Language Model Inference
Distributed Speculative Inference (DSI): Speculation Parallelism for Provably Faster Lossless Language Model Inference Open
This paper introduces distributed speculative inference (DSI), a novel inference algorithm that is provably faster than speculative inference (SI) [leviathan2023, chen2023, miao2024, sun2025, timor2025] and standard autoregressive inferenc…
View article: Dynamic Speculation Lookahead Accelerates Speculative Decoding of Large Language Models
Dynamic Speculation Lookahead Accelerates Speculative Decoding of Large Language Models Open
Speculative decoding is commonly used for reducing the inference latency of large language models. Its effectiveness depends highly on the speculation lookahead (SL)-the number of tokens generated by the draft model at each iteration. In t…
View article: Cross-Domain Aspect Extraction using Transformers Augmented with Knowledge Graphs
Cross-Domain Aspect Extraction using Transformers Augmented with Knowledge Graphs Open
The extraction of aspect terms is a critical step in fine-grained sentiment\nanalysis of text. Existing approaches for this task have yielded impressive\nresults when the training and testing data are from the same domain. However,\nthese …
View article: Efficient Few-Shot Learning Without Prompts
Efficient Few-Shot Learning Without Prompts Open
Recent few-shot methods, such as parameter-efficient fine-tuning (PEFT) and pattern exploiting training (PET), have achieved impressive results in label-scarce settings. However, they are difficult to employ since they are subject to high …
View article: TangoBERT: Reducing Inference Cost by using Cascaded Architecture
TangoBERT: Reducing Inference Cost by using Cascaded Architecture Open
The remarkable success of large transformer-based models such as BERT, RoBERTa and XLNet in many NLP tasks comes with a large increase in monetary and environmental cost due to their high computational load and energy consumption. In order…
View article: Opinion-based Relational Pivoting for Cross-domain Aspect Term Extraction
Opinion-based Relational Pivoting for Cross-domain Aspect Term Extraction Open
Domain adaptation methods often exploit domain-transferable input features, a.k.a. pivots. The task of Aspect and Opinion Term Extraction presents a special challenge for domain transfer: while opinion terms largely transfer across domains…
View article: InterpreT: An Interactive Visualization Tool for Interpreting Transformers
InterpreT: An Interactive Visualization Tool for Interpreting Transformers Open
Vasudev Lal, Arden Ma, Estelle Aflalo, Phillip Howard, Ana Simoes, Daniel Korat, Oren Pereg, Gadi Singer, Moshe Wasserblat. Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: System…
View article: Exploring the Boundaries of Low-Resource BERT Distillation
Exploring the Boundaries of Low-Resource BERT Distillation Open
In recent years, large pre-trained models have demonstrated state-of-the-art performance in many of NLP tasks. However, the deployment of these models on devices with limited resources is challenging due to the models’ large computational …
View article: Syntactically Aware Cross-Domain Aspect and Opinion Terms Extraction
Syntactically Aware Cross-Domain Aspect and Opinion Terms Extraction Open
A fundamental task of fine-grained sentiment analysis is aspect and opinion terms extraction. Supervised-learning approaches have shown good results for this task; however, they fail to scale across domains where labeled data is lacking. N…
View article: ABSApp: A Portable Weakly-Supervised Aspect-Based Sentiment Extraction System
ABSApp: A Portable Weakly-Supervised Aspect-Based Sentiment Extraction System Open
We present ABSApp, a portable system for weakly-supervised aspect-based sentiment extraction. The system is interpretable and user friendly and does not require labeled training data, hence can be rapidly and cost-effectively used across d…
View article: Multi-Context Term Embeddings: the Use Case of Corpus-based Term Set Expansion
Multi-Context Term Embeddings: the Use Case of Corpus-based Term Set Expansion Open
In this paper, we present a novel algorithm that combines multi-context term embeddings using a neural classifier and we test this approach on the use case of corpus-based term set expansion. In addition, we present a novel and unique data…
View article: ABSApp: A Portable Weakly-Supervised Aspect-Based Sentiment Extraction System
ABSApp: A Portable Weakly-Supervised Aspect-Based Sentiment Extraction System Open
Oren Pereg, Daniel Korat, Moshe Wasserblat, Jonathan Mamou, Ido Dagan. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-…
View article: Multi-Context Term Embeddings: the Use Case of Corpus-based Term Set Expansion
Multi-Context Term Embeddings: the Use Case of Corpus-based Term Set Expansion Open
In this paper, we present a novel algorithm that combines multi-context term embeddings using a neural classifier and we test this approach on the use case of corpus-based term set expansion. In addition, we present a novel and unique data…
View article: NLP Architect by Intel AI Lab
NLP Architect by Intel AI Lab Open
NLP Architect by Intel AI Lab: A Python library for exploring the state-of-the-art deep learning topologies and techniques for natural language processing and natural language understanding. Release v0.3 New Solution Topics and Trend Analy…
View article: Term Set Expansion based on Multi-Context Term Embeddings: an End-to-end Workflow
Term Set Expansion based on Multi-Context Term Embeddings: an End-to-end Workflow Open
We present SetExpander, a corpus-based system for expanding a seed set of terms into a more complete set of terms that belong to the same semantic class. SetExpander implements an iterative end-to end workflow for term set expansion. It en…
View article: Term Set Expansion based NLP Architect by Intel AI Lab
Term Set Expansion based NLP Architect by Intel AI Lab Open
We present SetExpander, a corpus-based system for expanding a seed set of terms into amore complete set of terms that belong to the same semantic class. SetExpander implements an iterative end-to-end workflow. It enables users to easily se…