Publications

2025

HELMET: How to Evaluate Long-Context Language Models Effectively and Thoroughly
Howard Yen, Tianyu Gao, Minmin Hou, Ke Ding, Daniel Fleischer, Peter Izsak, Moshe Wasserblat, Danqi Chen
Conference International Conference on Learning Representations (ICLR 2025)
Paper | OpenReview | Repo

2024

RAG Foundry: A Framework for Enhancing LLMs for Retrieval Augmented Generation
Daniel Fleischer, Moshe Berchansky, Moshe Wasserblat and Peter Izsak
Preprint Preprint
Paper | Repo
CoTAR: Chain-of-Thought Attribution Reasoning with Multi-level Granularity
Moshe Berchansky, Daniel Fleischer, Moshe Wasserblat, Peter Izsak
Conference The 2024 Conference on Empirical Methods in Natural Language Processing (EMNLP 2024) Findings
Paper | ACL Anthology

2023

Optimizing Retrieval-augmented Reader Models via Token Elimination
Moshe Berchansky, Peter Izsak, Avi Caciularu, Ido Dagan and Moshe Wasserblat
Conference The 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP 2023)
Paper

2022

Transformer Language Models without Positional Encodings Still Learn Positional Information
Adi Haviv, Ori Ram, Ofir Press, Peter Izsak and Omer Levy
Conference Findings of the Association for Computational Linguistics: EMNLP 2022
Paper

2021

How to Train BERT with an Academic Budget
Peter Izsak, Moshe Berchansky and Omer Levy
Conference The 2021 Conference on Empirical Methods in Natural Language Processing (EMNLP 2021)
Paper | Slides/Video | Repo

2020

Exploring the Boundaries of Low-Resource BERT Distillation
Moshe Wasserblat, Oren Pereg and Peter Izsak
Workshop Proceedings of SustaiNLP: Workshop on Simple and Efficient Natural Language Processing, 2020
Paper

2019

Training Compact Models for Low Resource Entity Tagging using Pre-trained Language Models
Peter Izsak, Shira Guskin and Moshe Wasserblat
Workshop 2019 Fifth Workshop on Energy Efficient Machine Learning and Cognitive Computing - NeurIPS Edition (EMC2-NIPS) (2019): 44-47.
Paper | Repo
Q8BERT: Quantized 8Bit BERT
Ofir Zafrir, Guy Boudoukh, Peter Izsak and Moshe Wasserblat
Workshop 2019 Fifth Workshop on Energy Efficient Machine Learning and Cognitive Computing - NeurIPS Edition (EMC2-NIPS) (2019): 36-39.
Paper | Repo

2018

Term Set Expansion based NLP Architect by Intel AI Lab
Jonathan Mamou, Oren Pereg, Moshe Wasserblat, Alon Eirew, Yael Green, Shira Guskin, Peter Izsak and Daniel Korat
Conference The 2018 Conference on Empirical Methods in Natural Language Processing (EMNLP 2018).
Paper | Repo
SetExpander: End-to-end Term Set Expansion Based on Multi-Context Term Embeddings
Jonathan Mamou, Oren Pereg, Moshe Wasserblat, Ido Dagan, Y. Goldberg, Alon Eirew, Yael Green, Shira Guskin, Peter Izsak and Daniel Korat
Conference Proceedings of the 27th International Conference on Computational Linguistics: System Demonstrations (COLING 2018)
Paper | Repo

2014

The search duel: a response to a strong ranker
Peter Izsak, Fiana Raiber, Oren Kurland and Moshe Tennenholtz
Conference Proceedings of the 37th international ACM SIGIR conference on Research & development in information retrieval (2014).
Paper

2013

Leveraging memory mirroring for transparent memory scale-out with zero-downtime failover of remote hosts
R. Tell, Peter Izsak, A. Shribman, Steve Walsh and B. Hudzia
Conference 2013 IEEE Symposium on Computers and Communications (ISCC) (2013)
Paper