I am a lecturer in computer science at the University of Birmingham. My research interests lay between formal methods and artificial intelligence. I study the question of how a computer can formally assure that an artificial intelligence (which is also a computer) acts safely. Here's my CV and here's a picture of me.

Contact

School of Computer Science, University of Birmingham
Edgbaston, Birmingham, B15 2TT, UK
m.giacobbe@bham.ac.uk

Selected Publications

Safe Artificial Intelligence

Verifying Reinforcement Learning up to Infinity. Edoardo Bacci, Mirco Gia­cob­be, David Parker. IJCAI, 2021. [paper]

Shielding Atari Games with Bounded Prescience. Mirco Giacobbe, Moham­madhosein Hasanbeig, Daniel Kroening, Hjalmar Wijk. AAMAS, 2021. [paper]

How Many Bits Does it Take to Quantize Your Neural Network? Mirco Gia­cob­be, Thomas A. Henzinger, and Mathias Lechner. TACAS, 2020.   [paper]

Proof Learning

Formal Synthesis of Lyapunov Neural Networks. Alessandro Abate, Daniele Ahmed, Mirco Giacobbe, and Andrea Peruffo. IEEE L-CSS, 2021.   [paper]

Learning Probabilistic Termination Proofs. Alessandro Abate, Mirco Gia­cob­be, and Diptarko Roy. CAV, 2021   [paper]

Neural Termination Analysis. Mirco Giacobbe, Daniel Kroening, and Julian Parsert. arXiv, 2021   [preprint]

Cyber-physical Systems

Continuous-Time Models for System Design and Analysis. Rajeev Alur, Mirco Giacobbe, Thomas A. Henzinger, Kim G. Larsen, Marius Mikucionis. Com­pu­ting and Soft­ware Sci­ence, 2019.   [paper]

Space-time Interpolants. Goran Frehse, Mirco Giacobbe, and Thomas A. Hen­zin­ger. CAV, 2018.   [paper]

Counterexample-guided Refinement of Template Polyhedra. Sergiy Bo­go­mo­lov, Goran Frehse, Mirco Giacobbe, and Thomas A. Henzinger. TACAS, 2017.   [paper] [tech. details]

Systems Biology

Model checking Gene Regulatory Networks (best paper award). Mirco Gia­cob­be, Ashutosh Gupta, Calin Č. Guet, Thomas A. Henzinger, Tiago Paixão, and Tatjana Petrov. TACAS, 2015.   [artifact] [paper] [extended paper]