On parle beaucoup de Thierry Breton ce matin. N'oublions pas Imran Ahmed (Center for Countering Digital Hate), Clare Melford (Global Disinformation Index), Anna-Lena von Hodenberg, fondatrice de HateAid et Josephine Ballon, de la même association.

I'm Sponsorship Co-Chair for #EACL2026 . EACL is a major NLP conference bringing together 2,000+ researchers and practitioners. Sponsorship boosts your organization’s visibility while supporting AI research. Sponsorship booklet: tinyurl.com/sponsor-eacl26

Finally done with my #ICLR2026 reviews! 🎉 Used the full rating scale this time: from 0 (a paper that should’ve been desk-rejected) to 10 (absolutely loved it). Average: 4.4.

The UTTER EU project is wrapping up! At NAVER LABS Europe, we’re proud to have contributed with models like mHuBERT & SPIRE as well as datasets such as ELITR-Bench for long-context LLM evaluation. Discover more in our latest post 👇 europe.naverlabs.com/blog/multimo...

Ma chorale se produit le WE des 27-28 Septembre à la salle Edmond Vigne de Fontaine dans l’agglomération grenobloise. Occasion de venir écouter et voir du chant 🎤mis en scène et chorégraphie 🕺. QR code sur le flyer.

Applications are now open. Deadline: October 12th !

Très heureux d’intervenir à l’école d’été #ETAL2025 à Roscoff (en terre bretonne 🌊) pour un cours d’introduction aux modèles multimodaux pour l’interaction orale. 📚 Matériel : github.com/besacier/ETA...

Je découvre à l'instant que la vuelta (tour d’espagne) passe devant chez moi aujourd’hui à seyssins, près de grenoble… lors d’une étape en provenance... d’italie 🤨

If you are leaving #acl2025 thinking "I wish for a week of time to think what's next", we have you covered: alps.imag.fr Announcing ALPS 2025, our 5th NLP winter school in 🇫🇷 We guarantee great scientific discussions, a fantastic scenery and great speakers:

🚀 A powerful and easy-to-build constrained speech LLM model. For multilingual SQA, we beat Microsoft-Phi and all unconstrained submissions on the SHORT track at IWSLT 2025 instruction-following track! Presented this Thursday at ACL 2025 by myself and @beomseok-lee.bsky.social 🧵👇

Sam Soul sera à Laval-en-Belledonne pour la fête de la musique le 21.6 🎷🎷🎷

As a scientist it should feel like an accomplishment when domain you’ve contributed to nears resolution after years of collective effort. So why many react with frustration refusing to let go of a field that’s essentially become engineering instead of moving on to new challenges?

🛥️ Hier soir, en marge du festival de Cannes, Lauren Sánchez a reçu un prix pour son «plaidoyer pour la justice climatique»… qu’elle est venue récupérer à bord du yacht de 127m de long qu’elle partage avec son fiancé, le patron d'Amazon Jeff Bezos. On vous en dit plus👇

Sam Soul fête ses 10 ans et vous invite à un apéro-concert le samedi 24.5 à 18h au lycée horticole (!) de St-Ismier 🌳 Entrée gratuite!

And we are done! Many tired eyes and lots of insights in today's project presentation (so much stuff done in 5 days!!) + insights from @dirkhovy.bsky.social #ALPS2025

#ALPS2025 lecture #5 was given by Alexandra Birch (University of Edinburgh). Title: “Reasoning and Knowledge in Multilingual Models”!!

#ALPS2025 lecture #4: Dzmitry Bahdanau (ServiceNow Research). Agents and Reasoning !

Yes, although it is April, #ALPS2025 is a truly WINTER school with snow activities etc. ❄️🎿🌨️

Wonderful lecture #3 at #ALPS2025 on Algorithms for text generation by François Yvon who introduced us to the large space of decoding algorithms!

We end day1 of #ALPS2025 with Marzieh Fadaee (Cohere For AI): “Languages as Windows to Human Knowledge: Building Better Language Models Through Global Understanding”

This year at #ALPS2025 we innovate with group mini projects proposed by organisers or participants themselves!

#ALPS2025 first talk: Kyunghyun Cho himself! “Causality gets help from deep learning “

#ALPS2025 day1 kicks off with a poster session!

Still possible to join this shared task !

ALPS 2025, the #NLP winter school in the Alps, kicks off in 2 days! 50 participants from around the world 🌍 7 top speakers, mini-projects, and 3 poster sessions ! Plus: snowy, screen-free fun in Aussois❄️ #ALPS2025 Web: https:/alps.imag.fr

Interesting to see the 20y-old Fisher dataset from 2004 making a strong comeback and being used to build dual-channel conversational speech models !

Join AutoMin 2025 - 3d Run of the Automatic Minuting Shared Task at #SIGDIAL2025 to push the limits of LLMs in 📌 Meeting Summarization 📌 Meeting Question Answering 📅 Workshop: Aug. 28, 2025 ufal.github.io/automin-2025/ #NLP #LLMs #AI #AutoMin

🤝 Supported by 10 beneficiaries: Uni Degli Studi Napoli Federico II, Uni de Extremadura, @csic.es, @uni-hamburg.de, Ensta Paris, @palrobotics.bsky.social, @naverlabseurope.bsky.social, Furhat Robotics, Uni de León, @sheffieldhallamuni.bsky.social ✍️ Info: euraxess.ec.europa.eu/jobs/317628 (3/3)

Welcome to NAVER LABS Europe on Bluesky! We chose 'LABS blue' to kickoff our posts & we'd appreciate it if you could share this one widely tinyurl.com/2nt9jdnv 😊 #hiring #jobs #internships

Sam Soul revient au local à St-Egreve le 15 Mars !

Worried that w/ rise of AI agents conducting multi-step research online (like OpenAI's Deep Research), human-made surveys will quickly lose visibility. We'll soon have a flood of non-opinionated AI-generated surveys that all look the same, marginalizing those created by humans...

C'est la dernière provocation du gouvernement : suspendre toutes les nouvelles missions de Service Civique tant que le budget 2025 n'aura pas été adopté. Ce chantage scandaleux maltraite encore une fois la jeunesse et le secteur associatif, déjà malmenés par 7 ans de macronisme.

🚨 Excited about ML/NLP and looking for a research internship on controlled text generation? Come work with us on advanced constraint processing in large language models at NAVER LABS Europe! ✨ Learn more and apply here: europe.naverlabs.com/job/2-7/

🚨🚨2025 Internship Call 🚨🚨 Come work with us on multimodal research for speech and text at NAVER LABS Europe 🤗 Apply here: europe.naverlabs.com/job/internsh... Please check the requirements on the next post 👇

I'll present two papers at the main #COLING2025 conference next week: (1) ELITR-Bench: A Meeting Assistant Benchmark for Long-Context Language Models (w/ T. Thonet and J. Rozen) - see arxiv.org/abs/2403.20262 and github.com/utter-projec...

🙌🏼 Excited to share our work on Speech Foundation Model for data crowdsourcing at COLING 2025 🙌🏼 Our co-author Laurent Besacier (@lbesacier.bsky.social) at NAVER LABS Europe will be presenting -- don't miss it. 👉🏼 Details: mt.fbk.eu/1-paper-acce...

Je viens de faire mon #eXit! L’exode de X est massif. Ne perdez pas un seul de vos abonnés. Grâce à #HelloQuitX j’ai inscrit 2221 nouveaux passagers pour partir vers #BlueSky. Retrouvez automatiquement vos communautés le #20Janvier via app.helloquitx.com !

The blog by session chairs & recordings of the #HRISymposium2024 sessions & panels are online 🫢 ! ⏯️Via blog: tinyurl.com/bxnafvu4 ⏯️Straight to playlist: tinyurl.com/ykdfnchh #hri #robotics #AI #NLProc

📣 Highlights of the 2nd International Symposium on Human Robot Interaction, organized earlier this month by NAVER Labs Europe! europe.naverlabs.com/blog/highlig...

Hi, so I've spent the past almost-decade studying research uses of public social media data, like e.g. ML researchers using content from Twitter, Reddit, and Mastodon. Anyway, buckle up this is about to be a VERY long thread with lots of thoughts and links to papers. 🧵

Hi, has anyone in my TL built something with the Moshi speech2speech model from Kyuteai or Mimi audio codec? The current release seems limited—no custom prompts, fine-tuning, voice changes, or delay control (making streaming TTS/ASR adaptations difficult). Any insights or workarounds?

Last month, I updated my 100-slide course on 'Automatic Speech Recognition: Introduction, Current Trends and Open Problems' for Univ Grenoble Alpes Master students! Dropped some Data2Vec and Transducer content, added fresh insights on Moshi S2S dialog model. Check it out:

Steve Johnson on using book text to build a text adventure in NotebookLM. “the game relies on three elements: the original text from my book; a large language model (… Gemini Pro 1.5); and a 400-word prompt that I wrote giving the model instructions on how to host the game” thelongcontext.com

A typical scenario: a rather solid 'dataset & resource' paper gets harshly criticized by an ICLR reviewer who gives it a score of 1, leading the authors to withdraw the paper... could affect ICLR topics' diversity...

I wanted to contribute to "Starter Pack Season" with one for Stanford NLP+HCI: go.bsky.app/VZBhuJ5 Here are some other great starter packs: - CSS: go.bsky.app/GoEyD7d + go.bsky.app/CYmRvcK - NLP: go.bsky.app/SngwGeS + go.bsky.app/JgneRQk - HCI: go.bsky.app/p3TLwt - Women in AI: go.bsky.app/LaGDpqg

Hi. Just landed here. Is there a starter pack with people working on #NLP and #LLMs to follow ? Tks.

I've started putting together a starter pack with people working on Speech Technology and Speech Science: go.bsky.app/BQ7mbkA (Self-)nominations welcome!

Talk: Speculations on Test-Time Scaling A tutorial on the technical aspects behind OpenAI's o1 and open research questions in this space. youtu.be/6PEJ96k1kiw Slides+bibliography: github.com/srush/awesom...

This is my first post here as I transition from Twitter/X, which has been ruined by EM. I'll stay there a bit longer during the transition. I share content on #NLP, #LLMs, #AI, and occasionally promote concerts in #Grenoble area from my soul band.