Michele Ciletti

📢 CLiC-it 2026 - Second Call for Papers is out! We're heading to Palermo (14-16 September 2026) for the very first time. Think Mediterranean sea views, Sicilian sunshine, and (yes) arancine 🌊☀️ ⏳ Deadline: 29 May 2026 🔗 Call: clic2026.unipa.it/call-for-paper… #NLProc #CLiCit2026

Advice for AI engineers 💡 You don't need a trillion-parameter model to change the world. A fine-tuned VLM like LFM2-VL + satellite imagery is enough. 10 real-world problems you can solve today 🧵↓

🏎️ Dataset: huggingface.co/datasets/MikCi… 📝 Cohere's model: cohere.com/transcribe

The dataset is available on Hugging Face. Feel free to take a look if you're testing audio models or if you're interested in studying how people communicate under pressure. (Or if you just want to see who talks the most on the radio - the data says Lewis and Max, unsurprisingly)

Fun weekend experiment: I built a dataset of ~15,000 Formula 1 team radio messages, with audio recordings and automatic transcriptions! 🏎️ (thread below) huggingface.co/datasets/MikCi…

Incredibly cool to see our community take Tiny Aya and make it their own. 🚀 From offline voice companions for kids 🎒 to tools for migrants 🌐 to pushing multilingual reasoning, safety, and cultural fluency 🦺. This is what Tiny Aya was built for: AI that runs anywhere, shaped by the people who use it. Learn more about Tiny Aya: cohere.com/blog/cohere-la…

This work is the direct outcome of my recent research stay at King's, funded by an H2IOSC Transnational Mobility Grant. Link to register in the post below! (2/2)

Thrilled to be part of the Computational Humanities seminar series at @KingsCollegeLon! On January 20th, I'll be presenting a talk on using LLMs for semantic annotation in historical and modern languages. (1/2)

Check out an amazing new article by @nadia_dileo on postdigital feminism, women's representation in video games, and intersectional experiences in gaming communities. Out now in Postdigital Science and Education! link.springer.com/article/10.100…

Merry Feminist Christmas! Another important article on postdiital feminism for the end of the year :) Link in first comment

At this point, papers testing whether AI can or cannot do something should try to test the strongest case, as well as a default. It is fine to say Llama 2 failed, but did a serious attempt to use GPT-5.1 Thinking in an agentic harness work? It would help better map the frontier.

[...] for perception verbs in Latin and Ancient Greek, while also testing these methods on Italian. The project will deliver new open data and a training resource on how to integrate LLMs in linguistic research. Many thanks to H2IOSC and its funders for supporting this work!

I'm glad to share that I have been awarded an H2IOSC Transnational Mobility Grant for a research stay at King’s College London! @kingsdh During the visit, I will work with @BarbaraMcGilli and @ImAndreaFarina to develop a semi-automatic semantic annotation pipeline [...]

Excited to speak at "AI and Digital Humanities: Methodological Approaches, Theories and Methods" at @unisiena on Nov. 5! I’ll discuss how hist. linguistics can partner with AI, presenting a study on Latin prosody. Check out the program and join us online: unisi.it/unisilife/even…

This week I had a lot of fun in Cagliari at #clicit2025! I presented a paper titled "Veras Audire Et Reddere Voces: A Corpus of Prosodically-Correct Latin Poetic Audio from Large-Language-Model TTS". Check out the pre-proceedings here: clic2025.unica.it/pre-proceeding…

🎉 Join us for "A Decade of Open Humanities Data", celebrating 10 years of the Journal of Open Humanities Data! 🗓️ 26 Sept 2025 🌐 Online & open to all Register by Sept 15 ⏰ openhumanitiesdata.metajnl.com/events/anniver… #JOHD10 #OpenHumanitiesData

Pictures from #clicit2025 social dinner!

Excited to be in Cagliari for #clicit2025! I'll be presenting a paper on text-to-speech LLMs and Latin prosody - check out the full program!

Great news from our #socialmedia team! 🚀 We’re excited to share that #JOHD is now on @LinkedIn! Managed by our @mik_cil 👏 Follow us to stay updated on: 🔹 New papers 🔹 Call for papers 🔹 News about #OpenData 👉 tinyurl.com/35cv9eu5 #OpenScience #HumanitiesData #LinkedIn