Log inSign up
user avatar
Mikel Artetxe
@artetxem
Co-founder @RekaAILabs and Honorary Researcher @Hitz_zentroa (University of the Basque Country) | Past: Research Scientist @AIatMeta (FAIR)
Basque Country
mikelartetxe.com
Joined July 2017
228
Following
6,999
Followers
  • Pinned
    user avatar
    Mikel Artetxe
    @artetxem
    Jul 9, 2025
    🚀 Excited to launch Reka Research, an AI agent that browses the web to answer your questions. Outperforming Gemini, GPT and Claude with web search! 🔥🔥🔥 Try it in our new playground (app.reka.ai/research) or get started for free with our API (docs.reka.ai/quick-start)
    user avatar
    Reka
    @RekaAILabs
    Jul 9, 2025
    🚀 Meet Reka Research––agentic AI that 🤔 thinks → 🔎 searches → ✏️ cites across the open web and private docs to answer your questions. 🥇 State-of-the-art performance, available now via our API and Playground!
    Image
    10K
  • user avatar
    Mikel Artetxe
    @artetxem
    Jul 29, 2020
    I'm happy to announce that I've successfully defended my thesis today, and I'll be joining @facebookai (FAIR) as a Research Scientist in September. I'm thankful for all the great people that I've met during my time in @IxaGroup and @txiomatika, and excited about my new adventure
    Image
  • user avatar
    Mikel Artetxe
    @artetxem
    Sep 30, 2019
    Who said that training GPT-2 or BERT was expensive? "We use 512 Nvidia V100 GPUs [...] Upon the submission of this paper, training has lasted for three months [...] and perplexity on the development set is still dropping."
    openreview.net
    Large-scale Pretraining for Neural Machine Translation with Tens of...
    In this paper, we investigate the problem of training neural machine translation (NMT) systems with a dataset of more than 40 billion bilingual sentence pairs, which is larger than the largest...
  • user avatar
    Mikel Artetxe
    @artetxem
    Dec 27, 2018
    Check out our new paper "Massively Multilingual Sentence Embeddings for Zero-Shot Cross-Lingual Transfer and Beyond" (w/ Holger Schwenk). New SOTA on cross-lingual transfer (XNLI, MLDoc) and bitext mining (BUCC) using a shared encoder for 93 languages! arxiv.org/abs/1812.10464
    Image
  • user avatar
    Mikel Artetxe
    @artetxem
    Mar 19, 2018
    We have just released UNdreaMT, a @PyTorch implementation of our Unsupervised Neural Machine Translation system. Train your own NMT model using nothing but monolingual corpora!
    Image
    GitHub - artetxem/undreamt: Unsupervised Neural Machine Translation
    From github.com
  • user avatar
    Mikel Artetxe
    @artetxem
    Apr 28, 2023
    I am fed up with all those "how good is ChatGPT at X" papers. Sharing some fresh ideas: - Evaluating Siri's ASR in noisy environments - How accurate is background removal in Zoom? - Google Translate in the biomedical domain - Can Gmail detect LLM-generated spam?
    62K
  • user avatar
    Mikel Artetxe
    @artetxem
    Mar 11, 2025
    Open sourcing our 21B reasoning model! Trained from scratch, strong numbers, Apache 2.0... Go build with it! 🔥🔥🔥 And even more excited to announce Nexus, our new platform to create AI workers, available in private preview. More coming soon!
    user avatar
    Reka
    @RekaAILabs
    Mar 11, 2025
    ⚡We are open sourcing Reka Flash 3, our new reasoning model that was trained from scratch. It achieves competitive performance with only 21B parameters. ⚡ Reka Flash 3 powers Nexus, our new enterprise intelligence platform. 🤖
    Image
    34K
  • user avatar
    Mikel Artetxe
    @artetxem
    Jan 9, 2019
    2018 was a productive year! I am 3rd in this ranking with 5 first author papers in top ML/NLP venues marekrei.com/blog/ml-and-nl… Unfortunately, the Spanish academic system is broken, and these publications may count close to zero if I decide to develop my career here. Thread 👇🏼
    Image
    marekrei.com
    ML and NLP Publications in 2018 - Marek Rei
    It is time for another yearly update of the publication statistics in Machine Learning and Natural Language Processing. The field has continued to grow very…
  • user avatar
    Mikel Artetxe
    @artetxem
    Nov 3, 2022
    📣 9 papers accepted at #emnlp2022 (7 main conference + 2 Findings) 🧵 with links to camera ready preprints 👇
  • user avatar
    Mikel Artetxe
    @artetxem
    Oct 28, 2019
    Check out our new paper "On the Cross-lingual Transferability of Monolingual Representations" (w/ @seb_ruder & @DaniYogatama) We challenge common beliefs of why mBERT works by showing that a monolingual BERT can also be transferred to new languages arxiv.org/abs/1910.11856
    Image
  • user avatar
    Mikel Artetxe
    @artetxem
    Apr 18, 2024
    Seriously?
    Image
    155K
  • user avatar
    Mikel Artetxe
    @artetxem
    Oct 24, 2018
    We are releasing Monoses, an open source implementation of our Unsupervised Statistical Machine Translation system. 26.2 BLEU in English-French WMT14 using nothing but monolingual corpora! - Paper: arxiv.org/abs/1809.01272 - Code: github.com/artetxem/monos…
  • user avatar
    Mikel Artetxe
    @artetxem
    Oct 31, 2017
    Check our new paper on "Unsupervised Neural Machine Translation". Training NMT models with only monolingual corpora! arxiv.org/abs/1710.11041
    Image
  • user avatar
    Mikel Artetxe
    @artetxem
    Mar 30, 2023
    📢 Life update 📢 After 2.5 wonderful years, I recently left FAIR to start a new adventure. It has been a privilege to be part of such an amazing team—I have learned a lot and had so much fun! I am super excited about what is coming next, and I hope to share more details soon!
    97K