Benjamin Marie,
researcher in Natural Language Processing (NLP)



I am a freelance research scientist in natural language processing. Previsouly, I was a researcher at 4i, in Sevilla (Spain), from March 2022 to July 2024. I mainly worked on improving multimodal dialogue engines using large language models. Before, I was a researcher at the Advanced Translation Technology Laboratory at NICT (Kyoto, Japan) from May 2016 to March 2022. My research focused on improving Machine Translation (MT) for low-resource language pairs, especially involving languages of East and South Asia.

Before joining NICT, I was a Ph.D. student at LIMSI-CNRS (Orsay, France), supervised by Aurélien Max and Anne Vilnat, also simultaneously engineer for the company Lingua-Et-Machina and sometimes teacher at Université Paris-Saclay.

Topics of interest : multimodal dialogue, low-resource neural MT, evaluation for MT, neural MT for user-generated texts

Grants/Fundings

Torres Quevedo grant (Spain). 3 years. Ended in July 2024.

NICT Tenure-track funding (Japan). 2 years. Ended in 2021.

JSPS (Japan Society for the Promotion of Science) grant for early-career scientists: Neural Machine Translation for User-Generated Contents. 2 years. Ended in 2022.

Committees

Best paper committees: ACL 2018 (Demo)

Paper reviewer: AACL (2020-), ACL (2017-), AAAI (2020-), COLING (2016-), COLM (2024-), EACL (2021-), EMNLP (2019*, 2017-), ICLR (2021-), IJCAI (2019-), IJCNLP (2017), LREC (2020,2018), NAACL (2016-), NeurIPS (2021-), TACL standing reviewer, JMLR, ACM TALLIP, IEEE/ACM TASLP, ACL Rolling Review (since 2021)
*: outstanding reviewer

Selected Publications
Full publication list here

2021 (First/Contact author only)

[pdf]
[bibtex]
[Annotations]

Marie, B., Fujita, A., Rubino, R. (2021). Scientific Credibility of Machine Translation Research: A Meta-Evaluation of 769 Papers. ACL 2021, online.
Outstanding Paper Award

2020 (First/Contact author only)

[pdf]
[bibtex]

Marie, B., Rubino, R., Fujita, A. (2020). Combination of Neural Machine Translation Systems at WMT20. In WMT20, online.
Ranked 1st (tied) for Ja-En, En-Iu, and Pl-En

[pdf]
[bibtex]

Marie, B., Fujita, A. (2020). Synthesizing Parallel Data of User-Generated Texts with Zero-Shot Neural Machine Translation. In TACL Vol. 8 (2020). Presented at ACL2021.

[pdf,bibtex]

Marie, B., Fujita, A. (2020). Iterative Training of Unsupervised Neural and Statistical Machine Translation Systems. In TALLIP Vol. 19 issue 5 (2020).

[pdf]
[bibtex]

Marie, B., Rubino, R., Fujita, A. (2020). Tagged Back-translation Revisited: Why Does It Really Work?. In ACL 2020, online.

2019

[pdf]
[bibtex]

Marie, B., Kaing, H., Mon, A.M., Ding, C., Fujita, A., Utiyama, M. and Sumita, E. (2019). Supervised and Unsupervised Machine Translation for Myanmar-English and Khmer-English. In WAT 2019, Hong Kong.
Ranked 1st for En->Km and Km->En.

[pdf]
[bibtex]

Marie, B., Sun, H., Wang, R., Chen, K., Fujita, A., Utiyama, M. and Sumita, E. (2019). NICT’s Unsupervised Neural and Statistical Machine Translation Systems for the WMT19 News Translation Task. In WMT19, Florence, Italy.
Ranked 1st.

[pdf]
[bibtex]

Marie, B., Dabre, R., and Fujita, A. (2019). NICT’s Machine Translation Systems for the WMT19 Similar Language Translation Task. In WMT19, Florence, Italy.

[pdf]
[bibtex]

Dabre, R., Chen, K., Marie, B., Wang, R., Fujita, A., Utiyama, M. and Sumita, E. (2019). NICT’s Supervised Neural Machine Translation Systems for the WMT19 News Translation Task. In WMT19, Florence, Italy.

[pdf]
[bibtex]

Marie, B. and Fujita, A. (2019). Unsupervised Joint Training of Bilingual Word Embeddings. In ACL 2019, Florence, Italy.

[pdf]
[bibtex]

Marie, B. and Fujita, A. (2019). Unsupervised Extraction of Partial Translations for Neural Machine Translation. In NAACL-HLT 2019, Minneapolis, USA.

2018

[pdf]

Marie, B., Fujita, A., Sumita, E. (2018). Combination of Statistical and Neural Machine Translation for Myanmar–English. In WAT 2018, Hong Kong.
Ranked 1st (BLEU) for My-En and En-My.

[pdf]
[bibtex]

Wang, R., Marie, B., Utiyama, M., Sumita, E. (2018). NICT's Corpus Filtering Systems for the WMT18 Parallel Corpus Filtering Task. In WMT18, Bruxelles, Belgium.

[pdf]
[bibtex]

Marie, B., Wang, R., Fujita, A., Utiyama, M., Sumita, E. (2018). NICT's Neural and Statistical Machine Translation Systems for the WMT18 News Translation Task. In WMT18, Bruxelles, Belgium.
Ranked 1st (BLEU) for Et-En, En-Et, En-Fi, and Fi-En.

[pdf]
[bibtex]

Marie, B. and Fujita, A. (2018). A Smorgasbord of Features to Combine Phrase-Based and Neural Machine Translation. In AMTA 2018, Boston, USA.

[pdf,bibtex]

Marie, B. and Fujita, A. (2018). Phrase Table Induction Using Monolingual Data for Low-Resource Statistical Machine Translation. In TALLIP Vol. 17 issue 3 (2018).

2017

[pdf]
[bibtex]

Marie, B. and Fujita, A. (2017). Phrase Table Induction Using In-Domain Monolingual Data for Domain Adaptation in Statistical Machine Translation. In TACL Vol. 5 (2017). Presented at ACL 2018

[pdf]
[bibtex]

Marie, B. and Fujita, A. (2017). Efficient Extraction of Pseudo-Parallel Sentences from Raw Monolingual Data Using Word Embeddings. In ACL 2017, Vancouver, Canada.

2015

[pdf]
[bibtex]

Marie, B. and Max, A. (2015). Touch-Based Pre-Post-Editing of Machine Translation Output. In EMNLP 2015, Lisbon, Portugal.

[pdf]
[poster]
[bibtex]

Marie, B., Allauzen, A., Burlot, F., Do, Q. K., Ive, J., Knyazeva, E., Labeau, M., Lavergne, T., Löser, K., Pécheux, N., Yvon, F. (2015). LIMSI@WMT'15: Translation Task. In WMT'15, Lisbon, Portugal.
Ranked 1st for En-Fr and Fr-En.

[pdf]
[poster]
[bibtex]

Marie, B. and Apidianaki, M. (2015). Alignment-based sense selection in METEOR and the RATATOUILLE recipe. In WMT'15, Lisbon, Portugal.
Ranked 1st for En-Fr and Fr-En.

[pdf]
[poster]
[bibtex]

Marie, B. and Max, A. (2015). Multi-Pass Decoding With Complex Feature Guidance for Statistical Machine Translation. In ACL-IJCNLP 2015, Beijing, China.

[pdf]
[poster]
[bibtex]

Apidianaki, M., Marie, B. (2015). METEOR-WSD: Improved Sense Matching in MT Evaluation. In SSST-9, Denver, US.

2014

[pdf]
[slides]
[bibtex]

Marie, B., Max, A. (2014). Confidence-based Rewriting of Machine Translation Output. In EMNLP 2014, Doha, Qatar.

[pdf]
[poster]
[bibtex]

Pécheux, N., Gong, L., Do, Q. K., Marie, B., Ivanishcheva, Y., Allauzen, A., Lavergne, T., Niehues, J., Max, A., Yvon, F. (2014). LIMSI @ WMT’14 Medical Translation Task. In WMT’14, Baltimore, US.

2013

[pdf]
[poster]
[bibtex]

Marie, B. and Max, A. (2013). A Study in Greedy Oracle Improvement of Translation Hypotheses. In IWSLT 13, Heidelberg, Germany.

Some Blog Posts (The Kaitchup - AI on a Budget)

Fast Speculative Decoding with Llama 3.2 and vLLM

DoRA vs. LoRA: Better and Faster than LoRA?

Train and Serve an AI Chatbot Based on Llama 3.2

QLoRA with AutoRound: Cheaper and Better LLM Fine-tuning on Your GPU

Run Llama 3.1 70B Instruct on Your GPU with ExLlamaV2 (2.2, 2.5, 3.0, and 4.0-bit)

Google's Gemma: Fine-tuning, Quantization, and Inference on Your Computer

Run Llama 3 70B on Your GPU with ExLlamaV2

SqueezeLLM: Better 3-bit and 4-bit Quantization for Large Language Models

QLoRA: Fine-Tune a Large Language Model on Your GPU

Generate Synthetic Data from Personas to Train AI Chatbots

Fine-Tuning Meta's Llama 3.2 1B & 3B Models on Budget GPUs

Multimodal RAG with ColPali and Qwen2-VL on Your Computer

How to Set Up a PEFT LoraConfig

Qwen2.5 QLoRA, LoRA, and Full Fine-tuning on Your Computer

Reports

2016

[pdf]

Ph.D. thesis: Complex Feature Guidance for Statistical Machine Translation (french)

2013

[pdf]

Project ANR TRACE report, part 5.2 (french)

2012

[pdf]

M.S. thesis: Improving Machine Translation Outputs by Greedy Search (french)