I can't believe this is really happening. WHAT A RACE, WHAT A RIDER!

Loving the new Kasabian album so far 💣. I hope they'll have a concert in Italy in the next few months

Can't wait to be in Brussels again :D

FOSDEM@fosdem.org · last wk.

Do LLMs Understand Their Users? An interesting paper introducing a dataset and benchmark for studying user interactions with LLMs, particularly how models can adapt their answers when user preferences are implicit or change over time.

What a trio!

First listen of the new Muse album. It's actually pretty good. I really loved the sound!

SIR LEWIS HAMILTON VINCE!

#Py4Ai 🔥🔥

We're excited to share that three papers from our lab have been accepted at #Interspeech2026 !! 🍾 #SpeechTranslation #SpeechAI #NLProc #FBK #Interspeech

Let's see if I have to skip the macOS update again this year or if they get rid of that ugly Liquid Glass

#PyconIT day 2 just started with a great opening keynote by @diegor.it about python performance

Four consecutive victories. No one had ever won four races after winning their first ever. 🇮🇹

⏰ 4 days left to apply! 🎓 2 PhD positions still open: 🎯 Human-Centred Evaluation Frameworks for Multilingual Technologies ✨ Multimedia Personalization with Multimodal Large Language Models 📅 Deadline: 15 May 2026 🔗 Full details: iecs.unitn.it/education/ad...

🤔 What Matters in Data for DPO? I asked myself this question a few days ago while trying to understand how to generate a dataset with preferences to run #DPO. This recent #NeurIPS paper answered some of my questions.

I'm getting used to this 🇮🇹

🎓 Come and join our group! 🎓 We offer 2 fully funded PhD positions: 🌍 Human-Centred Evaluation Frameworks for Multilingual Technologies (A6) 🤖 Multimedia Personalization with Multimodal Large Language Models (A7) ⏰ Deadline: 15 May 2026 🔗 Details: iecs.unitn.it/education/ad...

🎉 We have 2 papers accepted at the ACM FAccT conference 2026! 🎉 These papers are the result of a great collaboration with Xenia Heilmann and Mattia Cerrato from the University of Mainz. 🧵 1/5

🚀 New Shared Task: Model Compression for Machine Translation at hashtag#WMT2026 (co-located with hashtag#EMNLP2026)! 📅 Test data out on June 18th, submissions by July 2nd! Can you shrink an LLM and keep translation quality high? 🧠🔧 👉 www2.statmt.org/wmt26/model-... #NLP #ML #LLM #ModelCompression

I was looking for Artemis II wallpapers and found basicappleguy.com/haberdashery... for smartphones while this Reddit thread has some of then in 5120x2160 resolution www.reddit.com/r/Widescreen...

#Shrinking season 3 ended so perfectly that it could have been a series finale. If you’re already missing Bill Lawrence’s shows, give #Rooster on HBO Max a try.

an older man in a grey sweater says woof

ALT: an older man in a grey sweater says woof

media.tenor.com

🎉We’re happy to welcome our new research engineer, Yiu Chung Leung, joining us to work on multimodal LLMs with a focus on speech. Excited for what’s ahead! 🚀

At our last seminar, @maikezufle.bsky.social presented her work on Duplex Models, "Building Controllable Speech Systems"

Can we talk about the fact that we can just tell the computer > extract from 2026-04-08-17-15-58.mkv an audio file that is compatible with MacWhisper, ideally losslessly and it just runs ffmpeg for us now??

Not exactly the best advertisement for Claude Code and its ability to write its own code 1) github.com/anthropics/c... 2) github.com/anthropics/c...

[BUG] Claude Code login fails with OAuth timeout on Windows · Issue #44257 · anthropics/claude-code

Preflight Checklist I have searched existing issues and this hasn't been reported yet This is a single bug report (please file separate reports for different bugs) I am using the latest version of ...

github.com

With @malvinanissim.bsky.social and @vivipatti.bsky.social, we've been teaching ethics in NLP as a hands-on course across Groningen, Pavia & Turin. We wrote up the experience and received the ✨Best Paper Award✨ at #EACL2026's TeachNLP Workshop. Huge thanks to the organizers and all our students!

🎙️ Merve Noyan is keynoting #PyConIT2026! Open-source team at @HuggingFace, author of Vision Language Models (O'Reilly) "Open-source Multimodal AI": tools, state of the art, and how to get started. 🎟️ pycon.it/tickets #PyConIT #Python #AI

Core Dispatch #1 New releases are slated to drop on Monday: 3.13.13, 3.14.4, and 3.15.0a8. PEP 803 got accepted, so free-threaded builds will have a stable ABI. Ken Jin also wrote up what's going on with the JIT in 3.15. coredispatch.xyz/editions/1

Core Dispatch #1 — Core Dispatch

A regular digest of what's happening in CPython — from merged PRs and PEP decisions to community discussions and upcoming events.

coredispatch.xyz

Not a bad Sunday, starting with Kimi's victory in the Formula 1 Japanese Grand Prix and ending with Bez's victory in the United States GP.

Simply the Bez 🇮🇹

You did it again, Kimi! Great podium for Leclerc too, keeping Russell behind him.