Glad that my group, OmnAI Lab at Sapienza, got dealt a 🃏 Full House at the messy ICLR2026 with 5 over 5 papers accepted! 🃏✨

5 over 5 papers accepted at ICLR 2026

Glad that my group, OmnAI Lab at Sapienza, got dealt a 🃏 Full House at the messy ICLR2026 with 5 over 5 papers accepted! 🃏✨

Although this post looks like bragging, this was an amazing result achieved through hard, restless, and painful (!) work during 2024-2025!

Thank you all to my PhD students, colleagues, and other collaborators.

The “Winning Hand” of research:

♥️ A. R. Minut, H. Dewidar, I. Masi - Spilled Energy in Large Language Models

♠️ I. Maljkovic, M. R. Briglia, I. Masi, A. Cinà, F. Roli - Harnessing Hyperbolic Geometry for Harmful Prompt Detection and Sanitization - (led by SAIfer Lab)

♦️ M. R. Briglia, M. H. Mirza, G. Lisanti - I. Masi, Why Adversarially Train Diffusion Models?

♣️ A. D’Orazio, M. R. Briglia, D. Crisostomi, D. Loi, E. Rodolà, I. Masi - Implicit Inversion turns CLIP into a Decoder

🃏 Crisostomi, Zirilli, Gargiulo, Bucarelli, Scardapane, Silvestri, Masi, Rodolà - MASS: MoErging through Adaptive Subspace Selection - (led by GLADIA)

See you all in Rio de Janeiro, Brazil! 🇧🇷🌴

Acknowledgements