Marco Casadio

About

Marco Casadio is a researcher at the forefront of trustworthy artificial intelligence, with a primary focus on the verification and robustness of neural networks, particularly in Natural Language Processing (NLP). His major contribution is the development of ANTONIO, a systematic methodology for generating NLP benchmarks tailored to neural network verification—a notoriously difficult problem, as most verification techniques designed for numeric domains like computer vision fail for discrete, semantic NLP data. By identifying and addressing the technical reasons behind this gap, Casadio’s work provides a foundational framework for creating reliable benchmarks, enabling more rigorous testing of NLP models. His most-cited paper, "ANTONIO: Towards a Systematic Method of Generating NLP Benchmarks for Verification" (2023), has already garnered 3 citations, reflecting its emerging impact in the verification community. This achievement underscores his role in bridging the critical divide between verification theory and practical NLP applications, making his research essential for students and researchers aiming to build safer, more dependable language models.

Research Focus

Key Achievements

Top Papers

1. 1
   ANTONIO: Towards a Systematic Method of Generating NLP Benchmarks for Verification

   3 citations · 2023

Key Collaborators

Contact & Links