Julia Kreutzer

Julia Kreutzer

Dr. Julia Kreutzer

Former research assistant and PhD student at the Statistical NLP Group advised by Prof. Dr. Stefan Riezler. Now research scientist in the Google Translate team in Montreal, Canada.

All my code can be found on my Github page.

News

  • Paper accepted at EAMT. Coming soon.
  • Two papers accepted at the AfricaNLP workshop at ICLR, both in collaboration with the Masakhane team.
  • I will join the Translate team at Google Research Montreal as a Research Scientist.
  • Check out the Masakhane project: we’re building NMT for African languages (featuring Joey NMT).
  • We open-sourced Joey NMT, a minimalist NMT toolkit for NMT beginners based on Pytorch. The EMNLP’ 19 paper describes technical details, benchmark evaluations and a user study for code understanding.

Research Interests

  • Machine learning, in particular deep and reinforcement/bandit learning for NLP.
  • Machine translation and quality estimation.
  • Human-in-the-loop machine learning.

Curriculum Vitae

Education

2016-2020
PhD, Computational Linguistics; Heidelberg University.
2014-2016
Master of Arts, Computational Linguistics; Heidelberg University.

Thesis: Quality Estimation From Scratch

2011-2014
Bachelor of Arts, Computational Linguistics; Heidelberg University.

Thesis: Dimensionality Reduction in Semantic Vector Spaces Using a Derivational Resource

Experience

05/2019-08/2019
Research Intern; Google AI, Montreal.
03/2018-06/2018
Applied Science Intern; Amazon Development Center, Berlin. IWSLT 2018: “Learning to Segment Inputs for NMT Shows Preference for Character-Level Processing”
07/2017-10/2017
Research Intern; eBay, Aachen. NAACL 2018: “Can Neural Machine Translation be Improved with User Feedback?”
05/2016-10/2016
Research Intern; Unbabel, Lisbon. EMNLP 2017: “Learning What’s Easy: Fully Differentiable Neural Easy-First Decoders”

Teaching

Summer Term 2019
Co-instructor; undergraduate Software Project
Winter Term 2018
Instructor; undergraduate/graduate course “Neural Networks: Architectures and Applications for NLP”
Winter Term 2017
Instructor; undergraduate/graduate course “Neural Networks: Architectures and Applications for NLP”
Winter Term 2016
Instructor; undergraduate/graduate course “Neural Networks: Architectures and Applications for NLP”
Winter Term 2014
Teaching assistant; graduate course “Parallel Data Processing and Analysis” taught by Prof. Artur Andrzejak
Winter Term 2012
Teaching assistant; undergraduate course “Logical Foundations of Computational Linguistics” taught by Dr. Michael Herweg

Publications

2022

  1. Mayumi Ohta, Julia Kreutzer and Stefan Riezler
    JoeyS2T: Minimalistic Speech-to-Text Modeling with JoeyNMT
    Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing (EMNLP): System Demonstrations, Abu Dhabi, 2022
    pdf | code | bib 
    @inproceedings{joeys2t2022,
  author = {Ohta, Mayumi and Kreutzer, Julia and Riezler, Stefan},
  title = {Joey{S2T}: Minimalistic Speech-to-Text Modeling with {JoeyNMT}},
  journal = {Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing {(EMNLP)}: System Demonstrations},
  year = {2022},
  city = {Abu Dhabi},
  url = {https://arxiv.org/abs/2210.02545}
}

2020

  1. Julia Kreutzer, Stefan Riezler and Carolin Lawrence
    Learning from Human Feedback: Challenges for Real-World Reinforcement Learning in NLP
    “Challenges of Real-World RL” Workshop at the 34th Conference on Neural Information Processing Systems (NeurIPS), 2020
    pdf | bib 
    @article{kreutzer2020b,
  author = {Kreutzer, Julia and Riezler, Stefan and Lawrence, Carolin},
  year = {2020},
  title = {Learning from Human Feedback: Challenges for Real-World Reinforcement Learning in NLP},
  journal = {``Challenges of Real-World RL'' Workshop at the 34th Conference on Neural Information Processing Systems},
  journal-abbrev = {NeurIPS},
  url = {https://arxiv.org/abs/2011.02511}
}
  2. ∀, Iroro Orife, Julia Kreutzer, Blessing Sibanda, Daniel Whitenack, Kathleen Siminyu, Laura Martinus, Jamiil Toure Ali, Jade Abbott, Vukosi Marivate, Salomon Kabongo, Musie Meressa, Espoir Murhabazi, Orevaoghene Ahia, Elan van Biljon, Arshath Ramkilowan, Adewale Akinfaderin, Alp Öktem, Wole Akin, Ghollah Kioko, Kevin Degila, Herman Kamper, Bonaventure Dossou, Chris Emezue, Kelechi Ogueji and Abdallah Bashir
    Masakhane – Machine Translation For Africa
    “AfricaNLP” Workshop at the 8th International Conference on Learning Representations (ICLR), Addis Ababa, Ethiopia, 2020
    pdf | bib 
    @inproceedings{orife2020,
  author = {{$\forall$}, { } and Orife, Iroro and Kreutzer, Julia and Sibanda, Blessing and Whitenack, Daniel and Siminyu, Kathleen and Martinus, Laura and Ali, Jamiil Toure and Abbott, Jade and Marivate, Vukosi and Kabongo, Salomon and Meressa, Musie and Murhabazi, Espoir and Ahia, Orevaoghene and van Biljon, Elan and Ramkilowan, Arshath and Akinfaderin, Adewale and Öktem, Alp and Akin, Wole and Kioko, Ghollah and Degila, Kevin and Kamper, Herman and Dossou, Bonaventure and Emezue, Chris and Ogueji, Kelechi and Bashir, Abdallah},
  title = {Masakhane -- Machine Translation For Africa},
  journal = {``AfricaNLP'' Workshop at the 8th International Conference on Learning Representations},
  journal-abbrev = {ICLR},
  city = {Addis Ababa, Ethiopia},
  year = {2020},
  url = {https://arxiv.org/pdf/2003.11529}
}
  3. Elan van Biljon, Arnu Pretorius and Julia Kreutzer
    On Optimal Transformer Depth for Low-Resource Language Translation
    “AfricaNLP” Workshop at the 8th International Conference on Learning Representations (ICLR), Addis Ababa, Ethiopia, 2020
    pdf | bib 
    @article{biljon2020,
  author = {van Biljon, Elan and Pretorius, Arnu and Kreutzer, Julia},
  year = {2020},
  title = {On Optimal Transformer Depth for Low-Resource Language Translation},
  journal = {``AfricaNLP'' Workshop at the 8th International Conference on Learning Representations},
  journal-abbrev = {ICLR},
  city = {Addis Ababa, Ethiopia},
  url = {https://arxiv.org/abs/2004.04418}
}
  4. Julia Kreutzer, Nathaniel Berger and Stefan Riezler
    Correct Me If You Can: Learning from Error Corrections and Markings
    Proceedings of the 22nd Annual Conference of the European Association for Machine Translation (EAMT), Lisbon, Portugal, 2020
    pdf | bib 
    @article{kreutzer2020a,
  author = {Kreutzer, Julia and Berger, Nathaniel and Riezler, Stefan},
  year = {2020},
  title = {Correct Me If You Can: Learning from Error Corrections and Markings},
  journal = {Proceedings of the 22nd Annual Conference of the European Association for Machine Translation},
  journal-abbrev = {EAMT},
  city = {Lisbon, Portugal},
  url = {https://arxiv.org/abs/2004.11222}
}

2019

  1. Julia Kreutzer, Jasmijn Bastings and Stefan Riezler
    Joey NMT: A Minimalist NMT Toolkit for Novices
    Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP): System Demonstrations, Hong Kong, China, 2019
    pdf | code | bib 
    @inproceedings{joey2019,
  author = {Kreutzer, Julia and Bastings, Jasmijn and Riezler, Stefan},
  title = {Joey {NMT}: A Minimalist {NMT} Toolkit for Novices},
  journal = {Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing {(EMNLP-IJCNLP)}: System Demonstrations},
  year = {2019},
  city = {Hong Kong, China},
  url = {https://www.aclweb.org/anthology/D19-3019}
}
  2. Julia Kreutzer and Stefan Riezler
    Self-Regulated Interactive Sequence-to-Sequence Learning
    Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics (ACL), Florence, Italy, 2019
    pdf | bib 
    @inproceedings{kreutzer2019,
  author = {Kreutzer, Julia and Riezler, Stefan},
  title = {Self-Regulated Interactive Sequence-to-Sequence Learning},
  journal = {Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics},
  journal-abbrev = {ACL},
  year = {2019},
  city = {Florence},
  country = {Italy},
  url = {https://www.aclweb.org/anthology/P19-1029}
}

2018

  1. Julia Kreutzer and Artem Sokolov
    Learning to Segment Inputs for NMT Favors Character-Level Processing
    Proceedings of the International Workshop on Spoken Language Translation (IWSLT), Bruges, Belgium, 2018
    pdf | bib 
    @inproceedings{kreutzer2018c,
  author = {Kreutzer, Julia and Sokolov, Artem},
  title = {Learning to Segment Inputs for NMT Favors Character-Level Processing},
  journal = {Proceedings of the International Workshop on Spoken Language Translation},
  journal-abbrev = {IWSLT},
  year = {2018},
  city = {Bruges},
  country = {Belgium},
  url = {https://arxiv.org/abs/1810.01480}
}
  2. Julia Kreutzer, Joshua Uyheng and Stefan Riezler
    Reliability and Learnability of Human Bandit Feedback for Sequence-to-Sequence Reinforcement Learning
    Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (ACL), Melbourne, Australia, 2018
    pdf | bib 
    @inproceedings{kreutzer2018b,
  author = {Kreutzer, Julia and Uyheng, Joshua and Riezler, Stefan},
  title = {Reliability and Learnability of Human Bandit Feedback for Sequence-to-Sequence Reinforcement Learning},
  journal = {Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics},
  journal-abbrev = {ACL},
  year = {2018},
  city = {Melbourne},
  country = {Australia},
  url = {http://www.cl.uni-heidelberg.de/~riezler/publications/papers/ACL2018.pdf}
}
  3. Ryan Cotterell and Julia Kreutzer
    Explaining and Generalizing Back-Translation with Wake-Sleep
    arXiv preprint arXiv:1806.04402, 2018 (Preprint)
    pdf | bib 
    @article{cotterell2018,
  title = {Explaining and Generalizing Back-Translation with Wake-Sleep},
  author = {Cotterell, Ryan and Kreutzer, Julia},
  journal = {arXiv preprint arXiv:1806.04402},
  year = {2018},
  note = {Preprint},
  url = {https://arxiv.org/pdf/1806.04402.pdf}
}
  4. Tsz Kin Lam, Julia Kreutzer and Stefan Riezler
    A Reinforcement Learning Approach to Interactive-Predictive Neural Machine Translation
    Proceedings of the 21st Annual Conference of the European Association for Machine Translation (EAMT), Alicante, Spain, 2018
    pdf | bib 
    @inproceedings{lam2018,
  author = {Lam, Tsz Kin and Kreutzer, Julia and Riezler, Stefan},
  title = {A Reinforcement Learning Approach to Interactive-Predictive Neural Machine Translation},
  journal = {Proceedings of the 21st Annual Conference of the European Association for Machine Translation},
  journal-abbrev = {EAMT},
  year = {2018},
  city = {Alicante},
  country = {Spain},
  url = {http://www.cl.uni-heidelberg.de/~riezler/publications/papers/EAMT2018.pdf}
}
  5. Julia Kreutzer, Shahram Khadivi, Evgeny Matusov and Stefan Riezler
    Can Neural Machine Translation be Improved with User Feedback?
    Proceedings of the 16th Annual Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies - Industry Track (NAACL-HLT), New Orleans, LA, USA, 2018
    pdf | bib 
    @inproceedings{kreutzer2018a,
  author = {Kreutzer, Julia and Khadivi, Shahram and Matusov, Evgeny and Riezler, Stefan},
  title = {Can Neural Machine Translation be Improved with User Feedback?},
  journal = {Proceedings of the 16th Annual Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies - Industry Track},
  journal-abbrev = {NAACL-HLT},
  year = {2018},
  city = {New Orleans, LA},
  country = {USA},
  url = {http://www.cl.uni-heidelberg.de/~riezler/publications/papers/NAACL2018.pdf}
}

2017

  1. André F. T. Martins and Julia Kreutzer
    Learning What’s Easy: Fully Differentiable Neural Easy-First Taggers
    Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing (EMNLP), Copenhagen, Denmark, 2017
    pdf | NLP Highlights podcast | code | bib 
    @inproceedings{martins2017,
  title = {Learning What{'}s Easy: Fully Differentiable Neural Easy-First Taggers},
  author = {Martins, Andr{\'e} F. T. and Kreutzer, Julia},
  journal = {Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing},
  journal-abbrev = {EMNLP},
  year = {2017},
  city = {Copenhagen},
  country = {Denmark},
  url = {https://www.aclweb.org/anthology/D17-1036}
  doi = {10.18653/v1/D17-1036},
}
  2. Artem Sokolov, Julia Kreutzer, Kellen Sunderland, Pavel Danchenko, Witold Szymaniak, Hagen Fürstenau and Stefan Riezler
    A Shared Task on Bandit Learning for Machine Translation
    Proceedings of the Second Conference on Machine Translation (WMT), Copenhagen, Denmark, 2017
    pdf | shared task page | slides | bib 
    @inproceedings{sokolov2017,
  title = {A Shared Task on Bandit Learning for Machine Translation},
  author = {Sokolov, Artem and Kreutzer, Julia and Sunderland, Kellen and Danchenko, Pavel and Szymaniak, Witold and F{\"u}rstenau, Hagen and Riezler, Stefan},
  journal = {Proceedings of the Second Conference on Machine Translation},
  journal-abbrev = {WMT},
  year = {2017},
  city = {Copenhagen},
  country = {Denmark},
  url = {https://www.aclweb.org/anthology/W17-4756}
  doi = {10.18653/v1/W17-4756},
}
  3. Julia Kreutzer, Artem Sokolov and Stefan Riezler
    Bandit Structured Prediction for Neural Sequence-to-Sequence Learning
    Proceedings of the 55th Annual Meeeting of the Association for Computational Linguistics (ACL), Vancouver, Canada, 2017
    pdf | code | poster | bib 
    @inproceedings{kreutzer2017,
  author = {Kreutzer, Julia and Sokolov, Artem and Riezler, Stefan},
  title = {Bandit Structured Prediction for Neural Sequence-to-Sequence Learning},
  journal = {Proceedings of the 55th Annual Meeeting of the Association for Computational Linguistics},
  journal-abbrev = {ACL},
  year = {2017},
  city = {Vancouver},
  country = {Canada},
  url = {http://www.cl.uni-heidelberg.de/~riezler/publications/papers/ACL2017.pdf}
}

2016

  1. Artem Sokolov, Julia Kreutzer, Christopher Lo and Stefan Riezler
    Stochastic Structured Prediction under Bandit Feedback
    Advances in Neural Information Processing Systems (NIPS), Barcelona, Spain, 2016
    pdf | video | bib 
    @inproceedings{sokolov2016b,
  author = {Sokolov, Artem and Kreutzer, Julia and Lo, Christopher and Riezler, Stefan},
  title = {Stochastic Structured Prediction under Bandit Feedback},
  journal = {Advances in Neural Information Processing Systems},
  journal-abbrev = {NIPS},
  year = {2016},
  city = {Barcelona},
  country = {Spain},
  url = {http://www.cl.uni-heidelberg.de/~riezler/publications/papers/NIPS2016.pdf}
}
  2. Artem Sokolov, Julia Kreutzer, Christopher Lo and Stefan Riezler
    Learning Structured Predictors from Bandit Feedback for Interactive NLP
    Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (ACL), Berlin, Germany, 2016
    pdf | bib 
    @inproceedings{sokolov2016a,
  author = {Sokolov, Artem and Kreutzer, Julia and Lo, Christopher and Riezler, Stefan},
  title = {Learning Structured Predictors from Bandit Feedback for Interactive NLP},
  journal = {Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics},
  journal-abbrev = {ACL},
  year = {2016},
  city = {Berlin},
  country = {Germany},
  url = {http://www.cl.uni-heidelberg.de/~riezler/publications/papers/ACL2016.pdf}
}

2015

  1. Julia Kreutzer, Shigehiko Schamoni and Stefan Riezler
    QUality Estimation from ScraTCH (QUETCH): Deep Learning for Word-level Translation Quality Estimation
    Proceedings of the 10th Workshop on Machine Translation (WMT), Lisbon, Portugal, 2015
    pdf | bib 
    @inproceedings{kreutzer2015,
  author = {Kreutzer, Julia and Schamoni, Shigehiko and Riezler, Stefan},
  title = {QUality Estimation from ScraTCH (QUETCH): Deep Learning for Word-level Translation Quality Estimation},
  journal = {Proceedings of the 10th Workshop on Machine Translation},
  journal-abbrev = {WMT},
  year = {2015},
  city = {Lisbon},
  country = {Portugal},
  url = {http://www.cl.uni-heidelberg.de/~riezler/publications/papers/WMT2015.pdf}
}