What a day at our workshop on 'Risks and Unintended Harms of Generative AI Systems' at #ITADATA2025!

✨ Highlights:
- A keynote by Valerio Basile on perspectivism,
- Three paper presentations,
- A panel session with Amanda Cercas Curry, Alessandro Castelnovo, and Gianluca Bonifazi 💫

In Toronto for #KDD2025! 🇨🇦

Giving a hands-on tutorial on Concept-Based XAI with
@panisson.bsky.social
at 1 p.m. - materials: cxai-mechint-htutorial-kdd2025.github.io 💫
Also, Flavio Giobergia presents our paper on Interpretable Subgroup Drifts on Thursday! 🚀
doi.org/10.1145/3690...

Today, @riccardorenzulli.bsky.social held the seminar 'Learning Compositional and Interpretable Neural Representations' in our Explainable and Trustworthy AI course at PoliTO. 🙌
Thank you, Riccardo, for sharing your expertise with us!

Thank you, Prof. @jaom7.bsky.social (sqirllab.bsky.social‬), for sharing insights on Model Interpretation at our XAI seminar in the Explainable and Trustworthy AI course at PoliTo! 💡

He shared his work on coreset-based model interpretation and weight-based interpretability for neural networks.

Today, we held the first XAI seminar at the Explainable and Trustworthy AI course at PoliTO 🚀

Alan Perotti (CENTAI) presented his work on Operationalising XAI, describing practical challenges in XAI!
Thank you again, Alan! Great start to our series!

Last week, at our PoliTO&Amazon workshop, we presented the results of our collaboration, with great interventions from UniTO! 🤝
Exciting to see how strong the collaborations are! 🤝

Thanks to all presenters, @alkiskoudounas.bsky.social @gingerinai.bsky.social among others—excited for what’s next!✨

Had the pleasure of being part of Dr. Hamed Behzadi-Khormouji's PhD evaluation committee at the University of Antwerp
Well done today on your Pre-defence! You are almost there!

And when in Antwerp 🇧🇪... I presented my work at @sqirllab.bsky.social; thanks @jaom7.bsky.social for the invitation!